The effect of privatization on high-technology development

(1)

UNIVERSITY OF AMSTERDAM

DEPARTMENT ECONOMICS AND BUSINESS

Bachelor Thesis in Economics and Finance

The Effect of Privatization on

High-Technology Development

Author:

Supervisor:

Maarten Acda

Francisco Gomez Martinez

6126529

BSc Thesis E&F

29

th

of June 2015

Academic Year 2014/2015

(2)

2

Statement of Originality

This document is written by Student Maarten Cornelis Marinus Acda who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3

1. Introduction

An article published in The Economist on the 30th of July in 2013 looked at the changing global improvement in technology from the perspective of the country Estonia. A small Baltic state with around 1.3 million inhabitants. The article assumes that Estonia will become the world leader in technology. Intuitively not something you expect from a country that became independent from the Soviet Union in 1991. At that time half of its population had a telephone line, nowadays it has among the world’s zippiest broadband speeds. Kumar (2013). According to data of the World Bank: in 1995 the income out of high-technological exports per capita was far behind the world-wide average. Leading countries were countries like Japan and the United States of America. In 2013, Estonia was one of the world leaders in exports of high-technology export per capita and left superpowers like Japan and the United States far behind (World Bank 2013). The former prime-minister of Estonia, Mart Laar, explained this technological success through the implementation of flat income-tax, free trade, sound money and privatization in 1992. In the period 1988-2008 Estonia indeed had the highest amount of privatizations per capita in the world (International Finance

Corporation 1999). The current government aimed to transfer assets to companies and people with the incentives and skills to use them in an efficient way. In this thesis I am going to prove that privatization in a country has an effect on the share of high technology exports.

Innovation and development is essential for a country’s long-term economic growth and competitive advantage (Solow 1957). Stimulating innovation and development is very difficult, since stimulating, motivating, nurturing innovation and product development not only depend on internal structures of organizations. Technology capability misalignment and technological innovation and development exist outside the firm boundary as well.

(Mcelheran & Mcelheran 2013). The innovation and development process is long,

idiosyncratic and unpredictable, but also involves a high probability of failure (Hsu, Tian & Xu, 2014). Therefore, promoting innovation and development effectively requires well-functioning markets and the creation of a good climate for investors. Glisman and Horn (1988) argue that a stronger government intervention regarding technology production results in counter-productive effects. The incentives for investors should be restructured to assure a higher rate of capital formation. Despite the argument that strong government intervention and a low rate of capital formation holds back technological innovation and development in a country, empirical studies that link privatization and technological innovation are sparse.

(4)

4

Hence, the objective of this research is to provide cross-country evidence for the real effects of privatization on the economy from the perspective of technological innovation and development. Specifically I examine the impact of the amount of privatized companies and the size of proceeds for the privatized companies on the share of high-technology product exports. For privatization I use the following definition: The transfer of ownership and

control of State owned enterprise, all over the world (Bortolottie, Fantini & Siniscalco, 2001). High-technology products are defined as products with high R&D intensity, such as in

aerospace, computers, pharmaceuticals, scientific instruments, and electrical machinery (World Bank 2013).

A constraint for this research is that I do not control for import of high-technology. However, Cooper & Kleinschmidt (1985) explained that export routes are an important route to growth for firms, especially for high-technology and electronics firms. Therefore the firms who's sales mostly depend on export, the so called “world-marketers” perform best. Since in this research I want to test whether high-technology export outweigh the other exports of a country, through several privatizations, I will not control for the import of high-technology.

Since technological innovation causes growth and comparative advantages, it is essential for a country to know how it can stimulate this kind of innovation. Can a

government stimulate technological innovation by executing privatizations? And from which countries can I expect growth of their share of high-technology export? Moreover, when do you see an effect on your economy after privatizations?

This research provides some answers to these important questions, implementing an empirical analysis on a panel of 110 developing and less developed economies over the 1988-2008 period. My identification strategy is to use a panel-based fixed effects identification approach that signals the mechanism through which privatization has an effect on the share of export of high-technology products. I try to examine how much time it costs to see an effect of privatization and whether there is more effect from the size of proceeds of privatizations or the amount of privatizations in a country.

I collected the data about privatization proceedings and amounts form the International Finance Corporation (IFR). The data from the amount of exports of high-technology export is collected from the WorldBank (WB). The sample include all countries that executed privatizations In the period 1988-2008.

The research is organized as follows: In Section 2 I develop a testable hypothesis based on empirical findings. Section 3 includes a description of the data and provides a

(5)

5

summary of the statistics. Section 4 reports the empirical analysis and results. Finally Section 5 concludes and provides a discussion.

2. Hypotheses development

In this section I will develop a testable hypotheses by discussing the mechanism through which privatization affects the export of high-technology products and the accompanied innovation and development in the high-technology industry.

The development to a more high technology oriented economy has a huge impact on a country. There is a change in habits, cultural aspects, routines, regulation and education. In my research I measure this change in development of economies to a more high technology oriented economy through the export of high-technology products. There is an old debate that the development into a more high-technology oriented economy relies on the amount

patenting activities (Glisman & Horn, 1988). Reversely, Chuang (1998) claims that you cannot neglect the signal of the existing volume of demand for reversed engineered products or imitated technologies. He states that it is harder for countries to start from zero and to invent something genuinely new than add or modify features to existing products. Japan for example (note that Japan is not available in my sample) is a country that focused more on development than on research and patenting activities.

Besides Hall & Lerner (2010) provided evidence that knowledge assets which are output from research and development mainly produces intangible assets which are barely used to upgrade the efficiency of human capital.

There are several reasons why countries privatize their state-owned enterprises. In the empirical analysis I will control for the reasons why countries privatize. But why is there an effect of privatization activities on the export of high-technology products? In the

introduction I referred to Glisman and Horn (1988) who argue that government intervention regarding technology production results in a counter-productive effect. Important in the high-technology innovation and development process is an incentive for investors to be assured of a high rate of capital formation. A rationale for the reason why privatized companies are performing better than state-owned companies.

Another important effect of privatization is that privatized companies not only affect the internal incentives in firms for innovation and development activities of high-technology products. In markets wherein a substantial portion of government-owned assets were sold, the privatization decisions also resulted in opening up markets to firm entry (Aghion, Blundell,

(6)

6

Griffith, Howitt & Prantl 2009). Policies aiming at decreasing or removing product barriers to market entrance for incumbents are not sufficient to encourage growth of incumbent firms in all sectors of an economy. A decreased cautious attitude toward firm entry to product markets has several significant effects. While firm entrance to market is a major driver for economic growth, it is also a trigger to solve inefficiencies in product markets and affects the incentive for invention and development in incumbent firms.

Therefore privatizing activities result in an incentive for firms to innovate, develop and export high-technology products. In this context I will observe the effect on export of high-technology products, since you cannot neglect volume of demand for modified technology. Therefore I state the following empirical implication:

Hypothesis: Country’s privatization activities have a positive effect on the share of

high-technology exports.

The next section will describe how I bring these hypothesis to the data.

3. Data and summary statistics

3.1 Export of high-technology products

To implement the empirical analysis, I assembled a panel data set referring to a broad cross section of countries referring to their export of high-technology in current U.S. dollars for the 1988-2008 period. This data is based on the records of the World Bank (WB) and available at http://data.worldbank.org/indicator/TX.VAL.TECH.CD. For this research I selected a sample of 110 countries. It includes only developing countries as classified by the World Development Indicators of the World Bank. The dataset covers low-income, lower middle-income, and upper middle-income countries. High-income countries are not covered in the sample. Furthermore, developing countries where no privatizations are executed are not implemented in my sample. The selection of countries is applicable for my empirical

analysis. First, I am interested in countries who had a unavoidable amount of privatizations; second a dataset from centralized sources about privatization activities is only available for this list of countries.

The sample of high-technology exports in current U.S. dollars is an unbalanced panel. For all 2310 observations there are 777 values missing, but most countries have contiguous

(7)

7

observations. In case there was data missing, I noted a "missing value".

Due to data limitation I had to drop a few actively and less actively privatizing countries in the list: Afghanistan, Angola, Cape Verde, Cuba, Eritrea, Ethiopia, Guinea-Bissau, Kosovo, Lao PDR, Mauritania, Nepal, Sierra Leone and Uzbekistan. Most of the countries I had to drop for the reason that these countries are not reporting the export of high-technology products. However a missing value in high-high-technology exports doesn't mean that these countries are not engaged in high-technology export activities. Kosovo is an example of a country which is actively privatizing, but not reliable since the country exists only for a short period of time. In that case the data about the country is not reliable and I also had to drop Kosovo out of my sample.

The independent variable to measure whether a country develops itself toward a more high-technology oriented economy is constructed in the following way:

= share of high-technology exports.

My measure: share_tech_export*i,t is the percentage of high-technology exports of the total

amount of export for country i in year t. Table 1 provides a summary of the statistics.

Table 1:

In table 1 you see that in the sample the average share of high-technology export of total export is 2.87%. The minimum amount of percentage high-technology export is 9.08*10^-9% which is Belize in 2000 and the winner is the Philippines in 1999 with a 64.03% share of high-technology exports.

3.2 Explanatory variables

Privatization amounts and proceeds. I collect the annual privatization from data from the International Finance Corporation from the World Bank Group. It is a panel dataset referring to a broad cross section of developing countries for the period 1988 – 2008. It is available at:

share_tech~t 1533 .0287117 .0703409 9.08e-09 .6403415 Variable Obs Mean Std. Dev. Min Max

(8)

8

http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXTFINANCIALSECTOR/0,,con tentMDK:22936580~menuPK:7994350~pagePK:210058~piPK:210062~theSitePK:282885,0 0.html. It is a balanced dataset and the source reports privatization transaction of over U.S. $ 500.000 and the data is reported in millions of dollars. In the sample in the period 1988-2008, there are 7954 sales reported worth U.S. $ 590,000,000,000. It only includes transactions which generate proceeds or monetary receipts to the government (International Finance Corporation , 1999, 2008). Transactions that did not generate revenues for the government are excluded from the database. To find a quantitative indicator about the volume of state asset disposal by a given year I construct the two following variables: proceeds per capita in

year t and amounts per capita in year t (Bortolottie, Fantini & Siniscalco, 2001):

= proceeds per capita

My constructed variable proceeds per capita*i,t is the total amount of proceedings of

privatizations per capita in country i for year t

= amounts per capita

My constructed variable amounts_ per_capita*i,t is the total amount of privatizations per

capita in country i for year t. Because high-technology development takes time and the effect of the explanatory variables will be observed in future years, I have to build a delay in the privatization variables to test my empirical model. In my empirical test this delay will be noted as a lag. Where lag is the number of years prior to a certain amount of

high-technological development

In the dataset of International Finance Corporation transactions were classified into sector based on classifications made in the data sources. I neglect this information and take all privatizations in a certain year in account. In tables 2, 3, 4 and 5 you see the summary statistics.

Table 2: Summary proceeds in U.S.$

proceeds 2310 3.28e+08 2.29e+09 0 7.15e+10 Variable Obs Mean Std. Dev. Min Max

(9)

9

Table 3: Summary proceeds per capita in U.S.$

Table 4: Summary amount

Table 5: Summary amounts per capita

Appendix A reports a summary of the statistics of my independent variable and explanatory variable. It measures: total population in the period 1988-2008; the average share of high-technology exports in the period 1988-2008; the average amount of privatizations in the period 1988-2008 and the average proceeds in U.S.$ in the period 1988-2008.

3.3 Control variables

In my econometrics framework I need to control for explanatory variables that affect export of high-technology products and vary with country and year. The following variables are very conscious selected as control variables. One condition is that the control variables correlated with each other, since this can cause multicollinearity. I will control for four such variables: gross domestic product per capita; the political situation in a country expressed in the present form of government; industrial share of total value added and share of

expenditures on education from gross domestic product.

The control variable gross domestic product per capita is implemented due to the relation between value of shares privatized relative to GDP. A deep and liquid market allow the absorption of big issues. This makes it easier to privatize State owned enterprises. Furthermore, by exhibiting information in the announcement for a privatization, market liquidity facilitates monitoring and increases the market value of the public company. It allows the divesting government to raise more proceeds from the privatizations activities.

proceeds_p~a 2308 11.61536 41.39611 0 647.9593 Variable Obs Mean Std. Dev. Min Max

amount 2310 3.44329 13.73261 0 338 Variable Obs Mean Std. Dev. Min Max

amount_per~a 2308 5.31e-07 4.31e-06 0 .0001012 Variable Obs Mean Std. Dev. Min Max

(10)

10

(Bortolottie, Fantini & Siniscalco, 2001). The data is retrieved from the World Bank (World Bank 2014). Specifically my constructed variable gdp_ per_capita*i,t is gross domestic

product per capita in country i for year t. Similar to the explanatory variables the effect of gdp per capita will have an effect on the independent variable in future years. For that reason I have to build a delay in gdp_ per_capita*i,t to test my empirical model. In my empirical test

this delay will be noted as a lag. Where lag is the number of years prior to a certain amount of high-technological development. Table 6 reports a summary of the statistics.

Table 6: Summary GDP per capita

The panel is unbalanced. In case there was no information available about the gross domestic product for a certain year I dropped this observation. The reason observations are missing, is because these countries were not reporting that GDP or the reports were unreliable. Therefore missing observation doesn't mean that these countries do not have a gdp per capita.

Empirical research proves that privatization depends on a political dimension. The level of democracy is an important factor. Bortolottie, Fantini & Siniscalco (2001) proves that less established democracies with unstable political institutions appear barely able to set the sales of state owned enterprises in motion. The explanation for this is that privatization becomes less enforceable in less democratic settings since the incumbent government is forced to implement highly discounted fixed price offerings.

Since the re-election in highly democratic countries depends on the democratic support, a large scale privatization program can apply as a strategy for future support.

Through creating constituencies of voters interested in the maximization of the value of their assets, incumbent governments try to increase the democratic basis (Biais & Perotti 2002)

To construct a polity variable I retrieved data from the Polity IV Project of the Political Instability Task Force (PITF). Specifically I used the data in the column POLITY2 from this dataset. It contains dummies from -10 to 10 where the democracy indicator is an eleven point scale (0-10). The democratic scale is corrected for the autocratic scale, which contains dummies from -10 to 0. The democratic scale is corrected for autocracy since autocracies sharply restrict of suppress competitive political participation. Furthermore the dataset is unbalanced, but contiguous. In case there is a transformation to another polity or

gdp_per_ca~a 2199 2796.221 4387.552 97.15788 46402.64 Variable Obs Mean Std. Dev. Min Max

(11)

11

there is no incumbent government in case of a war, the dummies are labelled as unreliable and are not taken in account. (Polity IV Project 2014)

Specifically my constructed variable polity*i,t is the democratic scale corrected for the

autocratic scale in country i for year t. Table 6 reports a summary for polity for my sample:

Table 6: Summary polity:

Empirical literature provided systematic evidence that a certain level of polity has a direct effect on whether a country execute privatization. Since the explanatory variables need to be measured for their future effect on high-technology development in the future, I will also build in a delay for the control variable polity. In my empirical test this delay will be noted as a lag. Where lag is the number of years prior to a certain amount of high-technological development.

The following control variable is the industrial share of total value added per capita. The countries in the dataset are developed for heterogeneous degrees across different industries within a country. I control for the industrial share, because this reflects the

industry’s development and propensity to export high-technology products (Hsu, Tian & Xu, 2014). Specifically my constructed variable industry_value_added_per_capita*i,t is the

industrial share of total value added per capita in country i for year t. (In my empirical test I will not build a future effect on the high-technology development, since industry value added has a direct effect on the export of high-technology exports). The data is collected from the World Bank (World Bank 2014). The industry variable corresponds to ISIC divisions 10-45. These industrial Divisions are composed by the International Standard Industrial

Classification of All Economic Activities. There are 99 economical divisions in total. The missing data results from unreported observations.

Table 7 reports a summary for my selected sample (1988-2008). The numbers are reported in percentages.

polity 2088 2.163793 6.453846 -10 10 Variable Obs Mean Std. Dev. Min Max

(12)

12

Table 7: Summary industry value added per capita

The last control variable I will construct is the share of expenditures on education from gross domestic product. This variable has an effect on the share of exports of high-technology products through the mechanism that a bigger emphasis on education increases the value of human capital in a country (Griffith, Redding, & Van Reenen, 2004). A well-educated population is an important factor for the productiveness of the high-technology export industry. Expenditures on education is an important factor to increase the value of human capital in a country. The speed of development in the high-technology export sector products highly relies on the value of human capital working for a firm. It not only stimulates the development of high-technology products, but also effects innovations.

The data is assembled from the World Bank, (World Bank 2014) and is expressed in percentages. Specifically my constructed variable share_education_expenditure_gdp*i,t is the

expenditure on education divided by gross domestic product in country i for year t. The effects of expenditures on education will have a future effect. Like the explanatory variables and other control variables I will build in a delay for this control variable to compose a testable empirical test. This delay will be noted as a lag. Where lag is the number of years prior to a certain amount of high-technological development.

Table 8 reports a summary. The numbers reported in percentages. The missing observations are due to unreported expenditures.

Table 8: Summary of share of expenditure on education of gross domestic product

The actual sample size is determined by data availability. The final dataset is a unbalanced panel for the interval 1988-2008. Most countries have contiguous data, although the period countries started with reporting their data differ. Further all countries except Oman have holes in their series. In case there was no data available I didn’t take a certain

observation in account.

industry_v~a 2038 30.89968 10.56928 9.224226 77.41366 Variable Obs Mean Std. Dev. Min Max

share_educ~p 1108 4.158946 2.221387 .68942 44.33398 Variable Obs Mean Std. Dev. Min Max

(13)

13

4. Empirical analysis and results

In this section I present my empirical test and analysis and I will elaborate on my main findings and results. In the first section I will explain the econometric model I am going to test. The model will be exhibited and I will elaborate on how my data is applicable for my test. In the second section I examine how privatization and specific the proceeds per capita originating from privatization affect high-technology development. In the third section I examine how privatization and specific the amounts per capita affect high-technology development.

4.1 Econometric model

In the introduction I related to empirical literature that provided evidence that privatized companies are performing better than state-owned companies. In a seminal work, Glisman and Horn (1988) emphasize the importance for the incentive for investors to be assured for a high rate of capital formation. Privatization activities result in better performing markets wherein there is a higher potential for capital. Furthermore the interposition of a government regarding technology production results in a counter-productive effect. Inspired by their work, I propose the following model that extends their framework from a cross-country panel-dataset. By estimating the various variables in the model below, I examine the effect of privatization on high-technology development. Equation 1 represents the model.

Equation 1:

Technological development*i,t = β0 + β1 (Privatization*i,t+lag)

+ β2 (Gross Domestic Product*i,t+lag) + β3 (Polity*i,t+lag)

+ β4 (Industry Value*i,t) + β5 (Education*i,t+lag) + ε*i,t

Where Technological development*i,t+lag stand for share of high-technology exports.

Privatization is either amounts_ per_capita*i,t+lag and proceeds_per_ capita*i,t+lag . For the

regression I used fixed effects since I need to control for time varying country characteristics. Countries have time-invariant characteristics which should not correlate with other

(14)

14

whether random effect or fixed effects are preferred. The Probability was 0.0105 which is smaller than the probability of 0.05. So I control for fixed effects in the sample.

In the following section I will focus on the significance levels and signs of the various coefficients.

4.2 Dependence on privatization measured in proceeds per capita

To measure the effect of privatization I first elaborate on the output of the regression where proceeds_per_ capita*i,t+lag is the explanatory variable. I will focus on the sign and

significance level of β1. If a β1 has positive or negative sign and is significant it presumes that the size of proceeds for a privatized company after a certain amount of years exerts a positive or negative effect on the share of export of high-technology products. Appendix B reports the results form estimating equation 1, using proceeds_per_capita*i,t+lag as the variable

for Privatization*i,t+lag with a lag of one year.

In the regression with a lag of one year for proceeds_per_ capita*i,+t+lag, gdp_

per_capita*i,t+lag, polity*i,t+lag and share_education_expenditure_gdp*i,t+lag I only find that β2 is

positive and significant at the 5% level. After one year I cannot find a significant level for

proceeds_per_capita*i,t+lag.

For β2 I find the following coefficient estimates: 1.57*10^-6 (p-value = 0.010). Where gdp per capita is in U.S.$ and the share of high-technology export is in decimals. This estimate can be explained as followed: When gdp increases with U.S.$ 1,- in year zero, the share of high-technology increases with 0.000157% point in year 1. This can be illustrated as followed: In case a country can increases it’s gdp per capita with a U.S. $ 1000,- in year zero, it will cause an increase of 0.157 % point in share of high-technology exports in the

following year. The cross-country average share of high-technology exports for the period 1988-2008 is 2.87%.

Appendix C reports the results form estimating equation 1, using

proceeds_per_capita*i,t+lag as the variable for Privatization*i,t+lag with a lag of 9 years .The sign

of proceeds_per_ capita*i,+t+lag is not significant until nine years have passed. So after nine

years I find that the coefficient estimate of proceeds_per_ capita*i,t+lag, β1, is positive and

significant at the 5% level. I find that the coefficient estimate of β1 is 0.0000606 (p-value = 0.008). The following illustrates an explanation in other magnitudes: The proceeds per capita are in U.S.$ and the share of high-technology export is in decimals. If privatization yield U.S.$ 1,- than after nine years high-technology exports increase with 0.00606% point. A

(15)

15

more tangible example is the following: In a country with 1 million inhabitants, where the government sells a state-owned company worth U.S. $ 10,000,000,- in nine year the share of high-technological exports will increase 0.06% points due to the privatization proceedings. These preliminary findings appear to be consistent with my hypothesis.

4.3 Dependence on privatization measured in amounts per capita

The following privatization indicator I test is the amounts_ per_capita*i,t+lag. Again I

will focus on the sign and significance level of β1. If β1 has a positive or negative sign and is significant it presumes that the amount of privatized companies after a certain amount of years exerts a positive or negative effect on the share of export of high-technology products. Appendix D reports the results form estimating equation 1, using amounts_ per_capita*i,t+lag

as the variable for Privatization*i,t+lag with a lag of two year.

In the regression with a lag of two year for amounts_per_ capita*i,+t+lag, gdp_

per_capita*i,t+lag, polity*i,t+lag and share_education_expenditure_gdp*i,t+lag I find significant

coefficients at the 5% level for amounts per capita, gdp per capita and polity. The coefficient estimates of β1, β2 and β3 are -687.4463 (p-value = 0.005), 1.70*10^-6 (p-value = 0.013) and 0.00071 (p-value = 0.048), respectively.

The estimate for coefficient β1 will be explained here. Amounts per capita is expressed in one privatized company per capita and the share of high-technology export is expressed in decimals. The estimates of the coefficient can be illustrated in the following way: In a country with one million inhabitants, where in a certain year one state-owned company is privatized, the share high-technology exports will decrease with 0.0687 % point, two years after the executed privatization. The cross-country average share of

high-technology exports for the period 1988-2008 is 2.87%.

Like my findings for the effect of proceeds per capita, gdp per capita, β2, has an effect on the share of high-technology exports. The order of magnitude is equal. In case a country can increases its gdp per capita with a U.S. $ 1000,- in year zero, it will have cause an increase of 0.170 % point in share of high-technology exports in two years.

Unlike proceeds per capita, polity exerts an effect on the share of high-technology exports after two years. In case the democratic level increase by one on the 0-10 scale or the autocratic level increase by one on the -10-0 scale, the share of high-technology exports increases by 0.071% point in one year. Because, in case a lag of two year is applied, the

(16)

16

amount of privatization executed per capita has a negative effect on the share of high-technology exports, the preliminary findings appear to be inconsistent with my hypothesis.

Appendix E reports the results from estimating equation 1, using amounts_per_

capita*i,+t+lag as the variable for Privatization*i,t+lag, with a lag of three year for amounts_per_

capita*i,+t+lag, gdp_ per_capita*i,t+lag, polity*i,t+lag and share_education_expenditure_gdp*i,t+lag. I

find significant coefficients at the 5% level for amounts per capita and gdp per capita. The coefficient estimates of β1and β2 are 936.4636 (p-value = 0.000) and 2.61*10^-6 (p-value = 0.001), respectively.

Conversely to the lag of two years, I found a positive sign for the effect of

privatizations in amount per capita, β1, for a lag of three years. Coefficient β1stay positive and significant up until a lag of ten years. The estimate for the coefficient with a lag of three year can be explained on the following way: In a country with one million inhabitants, where in a certain year one state-owned company is privatized, the share high-technology exports will decrease with 0.0936 % point, two years after the executed privatization. The cross-country average share of high-technology exports for the period 1988-2008 is 2.87%. So for a country with 1 million inhabitants, which in a certain year privatize one company, you expect an increase of share of high-technology exports of 3.26 % (=0.0936 / 2.87 * 100).

Like my findings for a lag of two years, the effect of gdp per capita, β2, has an effect on the share of high-technology exports. The order of magnitude is equal. In case a country can increases its gdp per capita with a U.S. $ 1000,- in year zero, it will cause an increase of 0.261 % point in share of high-technology exports in two years.

This preliminary finding appear to be consistent with my hypothesis.

5. Conclusion and Discussion

This research presents a cross-country evidence on how privatization activities affect high-technological development. Using a dataset that includes data of 110 developing and less developed countries between 1988 and 2008 and a panel-based fixed effects identification approach, I identified that privatization activities exerts a positive effect on

high-technological development in the long run. I showed that the amount of privatizations have a stronger effect on high-technological development compared to the proceeds for certain privatization deals. Additionally the effect of the amount of privatizations on high-technology development is exerted with a shorter lag, compared to the proceeds resulting in a

(17)

17

on high-technological development shortly after privatization activities. Although this effect was positively corrected in the years thereafter, this can be interesting for further research as well as the different lags in effect of results of proceeding and amounts. Referring to the introduction I can conclude that Mart Laar, the former president of Estonia was right, privatization has an effect on the high-technological development of a country. My research offers new insight on the effect of privatization on high-technological development.

References

Aghion, P., Blundell, R., Griffith, R., Howitt, P., & Prantl, S. (2009). The effects of entry on incumbent innovation and productivity. The Review of Economics and Statistics, 91(1), 20-32.

Biais, B., & Perotti, E. (2002). Machiavellian privatization. The American Economic Review, 92(1), 240-258.

Bortolotti, B., Fantini, M., Siniscalco, D. (2001) Privatisation around the world: New evidence form Panel Data. Journal of Public Economics(600)

Chuang, Y. C. (1998). Learning by doing, the technology gap, and growth. International Economic Review, 697-721.

Cooper R. G., Kleinschmidt, E. J.(1985) The Impact of Export Strategy on Export Sales Performance. Journal of International Business Studies(16)(1)(37-55)

Glismann, H. H., Horn, E., (1988). Comparative Invention Performance of Major Industrial Countries: Patterns and Explanations. Management Science,(34)(10), 1169 – 1187. Griffith, R., Redding, S., & Van Reenen, J. (2004). Mapping the two faces of R&D:

productivity growth in a panel of OECD industries. Review of Economics and Statistics, 86(4), 883-895.

Hsu, P., Tian, X., Xu, Y., (2014) Financial development and innovation: Cross-country evidence. Journal of Financial Economics(112),(1),116-135

International Finance Corporation (1999). PrivatisationData88_99. Retrieved from:

http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXTFINANCIALSECTO R/0,,contentMDK:22936580~menuPK:7994350~pagePK:210058~piPK:210062~theS itePK:282885,00.html

International Finance Corporation (2008). PrivatisationData88_99. Retrieved from:

(18)

18

R/0,,contentMDK:22936580~menuPK:7994350~pagePK:210058~piPK:210062~theS itePK:282885,00.html

Hall, B. H., & Lerner, J. (2010). The financing of R&D and innovation. Handbook of the Economics of Innovation, 1, 609-639.

Kumar, A. (2013). How did Estonia become a leader in technology? [Blog post]. Retrieved from: http://www.economist.com/blogs/economist-explains/2013/07/economist-explains-21

Mcelheran, K., Mcelheran, B. (2013). Do Market Leaders Lead in Business Process

Innovation? The Case(s) of E-Business Adoption. Management Science, January, (1-32)

Polity IV Project. (2014). P4v2014. Retrieved from: http://www.systemicpeace.org/polity/polity4.htm

World Bank. (2013). High-technology exports (current US$). Retrieved from: http://data.worldbank.org/indicator/TX.VAL.TECH.CD?display=graph World Bank (2013). Population, total. Retrieved from:

http://data.worldbank.org/indicator/SP.POP.TOTL

World Bank (2014) Government expenditure on education, total (% of GDP). Retrieved from: http://data.worldbank.org/indicator/SE.XPD.TOTL.GD.ZS/countries?page=3 World Bank (2014). GDP per capita (current US$). Retrieved from:

http://data.worldbank.org/indicator/NY.GDP.PCAP.CD

World Bank (2014). Industry, value added (% of GDP). Retrieved from: http://data.worldbank.org/indicator/NV.IND.TOTL.ZS

(19)

19

Appendix A: Country summary for the period 1988 - 2008

country average total population in millions average share of high-technology exports average amount of privatizations average proceeds privatizations in millions Albania 3.1 0.27% 2.71 $ 24.25 Algeria 30.6 0.03% 0.48 $ 75.98 Argentina 35.9 1.86% 6.57 $ 2,123.00 Armenia 3.2 0.67% 5.57 $ 19.77 Azerbaijan 7.9 0.32% 1.52 $ 9.61 Bahrain 0.7 0.09% 0.05 $ 0.49 Bangladesh 126.7 0.19% 1.81 $ 7.42 Barbados 0.3 1.42% 0.29 $ 2.43 Belarus 10.0 1.80% 3.05 $ 16.09 Belize 0.2 0.03% 0.33 $ 6.74 Benin 6.7 0.01% 0.71 $ 5.97 Bhutan 0.6 0.87% 0.05 $ 0.86 Bolivia 8.2 3.45% 4.24 $ 160.60

Bosnia and Herzegovina 3.9 1.14% 0.52 $ 45.93

Brazil 169.1 4.58% 9.10 $ 4,201.00

Bulgaria 8.2 2.21% 17.90 $ 322.00

Burkina Faso 11.2 0.43% 0.43 $ 15.29

Burundi 6.7 0.19% 0.29 $ 0.18

Cameroon 15.2 0.11% 0.48 $ 18.42

Central African Republic 3.5 0.07% 0.05 $ 0.57

Chile 15.0 0.48% 1.81 $ 192.60 China 1230.7 15.77% 15.52 $ 9,111.00 Colombia 38.6 1.68% 1.95 $ 554.10 Congo. Rep. 3.0 0.31% 0.05 $ 0.10 Costa Rica 3.7 14.95% 0.43 $ 5.38 Cote d'Ivoire 15.1 0.99% 2.48 $ 28.48 Croatia 4.5 3.68% 1.57 $ 225.90 Czech Republic 10.3 8.62% 5.86 $ 868.80 Dominican Republic 8.4 1.90% 0.19 $ 30.64 Ecuador 12.1 0.37% 0.67 $ 42.83

Egypt. Arab Rep. 64.4 0.05% 4.67 $ 198.70

El Salvador 5.8 1.56% 0.38 $ 65.00 Equatorial Guinea 0.5 . 0.05 $ 0.10 Estonia 1.4 8.76% 20.52 $ 48.46 Fiji 0.8 0.39% 0.14 $ 1.93 Gabon 1.2 0.63% 0.19 $ 4.17 Gambia. The 1.2 0.04% 0.00 $ -Georgia 4.6 2.53% 6.14 $ 86.67 Ghana 18.2 0.18% 4.81 $ 98.84 Grenada 0.1 1.23% 0.05 $ 0.29 Guatemala 10.8 1.52% 0.24 $ 64.98 Guinea 8.2 0.02% 0.10 $ 3.95 Guyana 0.7 0.18% 0.24 $ 3.23 Haiti 8.3 1.37% 0.05 $ 0.79 Honduras 6.0 0.18% 1.95 $ 8.53 Hungary 10.3 13.53% 42.14 $ 1,034.00 India 1006.8 2.91% 6.81 $ 885.50 Indonesia 203.1 4.34% 1.81 $ 431.40 Iran 63.6 0.14% 0.05 $ 16.67 Iraq 22.6 0.00% 0.05 $ 59.52

(20)

20

country average total population in millions average share of high-technology exports average amount of privatizations average proceeds privatizations in millions Kazakhstan 15.6 1.55% 1.29 $ 448.30 Kenya 29.9 0.63% 4.14 $ 75.42 Kyrgyz Republic 4.8 1.63% 1.19 $ 6.76 Latvia 2.4 1.87% 1.81 $ 68.37 Lebanon 3.3 0.48% 0.10 $ 17.05 Lesotho 1.8 0.21% 0.14 $ 1.57 Lithuania 3.5 2.33% 5.90 $ 137.60 Macedonia. FYR 2.0 0.83% 26.14 $ 68.43 Madagascar 15.0 0.34% 0.19 $ 2.82 Malawi 11.1 0.12% 1.05 $ 2.73 Malaysia 22.3 30.43% 2.33 $ 592.40 Mali 10.0 0.15% 0.43 $ 4.41 Mauritius 1.2 1.22% 0.05 $ 12.43 Mexico 99.8 11.47% 13.14 $ 1,949.00 Moldova 3.6 0.95% 0.76 $ 2.27 Montenegro 0.6 0.20% 0.05 $ 1.02 Morocco 27.7 2.99% 3.95 $ 524.70 Mozambique 17.5 0.30% 7.38 $ 13.92 Namibia 1.8 1.72% 0.05 $ 0.40 Nicaragua 4.9 0.27% 3.90 $ 16.40 Niger 10.5 0.44% 0.24 $ 4.89 Nigeria 118.4 0.03% 7.71 $ 345.00 Oman 2.2 0.60% 0.38 $ 43.43 Pakistan 136.4 0.42% 8.29 $ 454.70 Panama 2.9 0.02% 0.86 $ 76.16

Papua New Guinea 5.2 1.53% 0.05 $ 10.65

Paraguay 5.1 0.15% 0.10 $ 2.00 Peru 25.0 0.41% 8.86 $ 455.30 Philippines 74.5 40.52% 5.24 $ 389.80 Poland 38.3 1.76% 23.57 $ 1,346.00 Romania 22.2 2.08% 4.29 $ 629.40 Russian Federation 146.3 2.35% 5.67 $ 2,894.00 Rwanda 7.8 0.79% 0.19 $ 6.09

Sao Tome and Principe 0.1 0.02% 0.00 $

-Senegal 9.5 1.07% 0.48 $ 32.66 Serbia 7.5 1.97% 4.76 $ 297.30 Slovakia 5.4 2.66% 1.57 $ 332.40 Slovenia 2.0 3.61% 0.62 $ 64.04 South Africa 41.8 2.16% 1.19 $ 237.70 Sri Lanka 18.5 0.90% 4.81 $ 42.42 Sudan 26.4 0.05% 0.14 $ 11.71 Syria 15.6 0.11% 0.10 $ 3.33 Tajikistan 6.1 4.37% 0.05 $ 0.06 Tanzania 32.6 0.13% 3.71 $ 21.48 Thailand 61.1 15.98% 1.24 $ 287.40 Togo 4.7 0.09% 0.48 $ 1.85

(21)

21

country average total population in millions average share of high-technology exports average amount of privatizations average proceeds privatizations in millions Uganda 23.3 0.56% 3.19 $ 15.92 Ukraine 49.7 2.43% 1.76 $ 283.60

United Arab Emirates 3.1 0.03% 0.05 $ 9.05

Uruguay 3.2 0.61% 0.38 $ 23.63 Venezuela 23.5 0.29% 2.90 $ 289.10 Vietnam 75.0 2.24% 6.33 $ 43.70 Yemen 16.4 0.00% 0.14 $ 11.14 Zambia 9.7 0.34% 3.48 $ 39.38 Zimbabwe 11.9 1.27% 0.67 $ 16.62

(22)

22

F test that all u_i=0: F(93, 711) = 77.64 Prob > F = 0.0000 rho .87349852 (fraction of variance due to u_i)

sigma_e .02590715 sigma_u .06807738 _cons .0122198 .0103148 1.18 0.237 -.0080312 .0324708 L1. .0023886 .0008371 2.85 0.004 .0007451 .004032 share_educa~p industry_va~a .0003082 .0003118 0.99 0.323 -.000304 .0009205 L1. .000669 .0004061 1.65 0.100 -.0001282 .0014662 polity

L1. 1.57e-06 6.10e-07 2.57 0.010 3.68e-07 2.76e-06 gdp_per_cap~a

L1. -.0000176 .0000181 -0.97 0.331 -.0000532 .0000179 proceeds_pe~a

share_tech_~t Coef. Std. Err. t P>|t| [95% Conf. Interval] corr(u_i, Xb) = 0.1083 Prob > F = 0.0005 F(5,711) = 4.50 overall = 0.0454 max = 18 between = 0.0286 avg = 8.6 R-sq: within = 0.0307 Obs per group: min = 1 Group variable: country_id Number of groups = 94 Fixed-effects (within) regression Number of obs = 810

(23)

23

Appendix C:

sigma_e .01782878 sigma_u .07690016 _cons .0523319 .0115438 4.53 0.000 .0296181 .0750456 L9. -.0011018 .0014339 -0.77 0.443 -.0039232 .0017195 share_educa~p industry_va~a -.0002704 .000334 -0.81 0.419 -.0009275 .0003867 L9. -.0004245 .0004485 -0.95 0.345 -.001307 .000458 polity

L9. 3.88e-07 1.79e-06 0.22 0.829 -3.13e-06 3.91e-06 gdp_per_cap~a

L9. .0000606 .0000226 2.68 0.008 .0000161 .000105 proceeds_pe~a

share_tech_~t Coef. Std. Err. t P>|t| [95% Conf. Interval] corr(u_i, Xb) = -0.1230 Prob > F = 0.0967 F(5,311) = 1.88 overall = 0.0047 max = 12 between = 0.0101 avg = 4.6 R-sq: within = 0.0294 Obs per group: min = 1 Group variable: country_id Number of groups = 87 Fixed-effects (within) regression Number of obs = 403

(24)

24

Appendix D:

sigma_e .02233925 sigma_u .0711158 _cons .0284676 .0088847 3.20 0.001 .011023 .0459121 L2. .0011301 .0006879 1.64 0.101 -.0002206 .0024808 share_educa~p industry_va~a -.0000653 .0002651 -0.25 0.806 -.0005857 .0004552 L2. .00071 .0003577 1.99 0.048 7.72e-06 .0014122 polity

L2. -687.4463 246.2872 -2.79 0.005 -1171.013 -203.8791 amount_per_~a

(25)

25

Appendix E:

sigma_e .02105354 sigma_u .07463702 _cons .0324498 .0086963 3.73 0.000 .0153732 .0495264 L3. .0000138 .0006466 0.02 0.983 -.0012559 .0012836 share_educa~p industry_va~a -.0000871 .000259 -0.34 0.737 -.0005957 .0004215 L3. .0005341 .000355 1.50 0.133 -.0001629 .0012311 polity

L3. 936.4636 230.7252 4.06 0.000 483.3978 1389.529 amount_per_~a