This research is intended to be done because of its applicable and useful results. To make it crystal clear, the results can be applied to the growth of high-tech exports in selected countries. It is descriptive-analytical in terms of nature since it describes and analyzes the relationship between the variables, using secondary statistics without interference and manipulation. We use multivariate regression analysis (the core of econometric studies), panel data approach, GMM, and Stata software to estimate the effect of national brand on high-tech exports. The statistical sample of this paper consists of 14 developing countries[6] (sample group) and 12 emerging developed countries[7] (control group) whose data are available over the period of 2011–2018.
4-1. MODEL PRESENTATION
The model used in this paper is a panel data type that can provide a more efficient estimation due to the limitation of the variance heterogeneity problem, decrease in the coherence between variables, and increase in the degree of freedom over cross-sectional and time series data (Baltaghi, 2005). The panel data consists of both static and dynamic types. The model of this paper is dynamic in which the dependent variable’s lag appears as the explanatory variable on the right side to clear the relationships between the variables (Arellano and Bond, 1991). This is because many of the economic variables, including high-tech exports, are naturally dynamic, and their performance of the previous period can be extended to the next period.
However, in the dynamic panel model, due to the addition of the interrupt dependent variable, it is not possible to use conventional estimation methods such as ordinary least squares (OLS), least-squares dummy variables (LSDV), and generalized least squares (GLS). Indeed, the perturbation component correlates with the interrupt-dependent variable and the estimation results are distorted. Therefore, Arlano and Bond (1991) proposed an estimator called Generalized Method of Moments (GMM) which, not only solves correlation problem between the independent variable and the perturbation component, also it eliminates variables’ indigenousness and variance heterogeneity of the model. This estimator works on both fixed and random effects and does not require the Hausman test because in dynamic panel models, there is a relationship between error term and explanatory variables (Hayashi, 2000). In addition, this method is often used when the number of cross-sectional variables (N) is greater than the number of times and years (T) (T <N), similar to the condition, which this study has been carried out.
With these explanations, the sub-form of the high-tech export equation inspired by the Sezer (2018), Bayraktutan and Bidirdi (2018), Kabaklarli et al. (2017), Mehrara et al. (2017), Alagöz et al. (2016), Sandu and Ciocanel (2014), Gökmen and Turen (2013), and Tebaldi (2011) works are presented as follows.
In Equation 1, t denotes time, i refers to the since the logarithmic form represents the percentage change of the dependent variable vs the percentage change of the explanatory variable, all variables are logarithmic to simplify coefficients interpretation, and stands for model error term that includes country-specific fixed effects (vi) and error residual term (eit). In addition, all variables are logarithmically assigned to simply interpret the coefficients since the logarithmic form represents the percentage change of the dependent variable vs the percentage change of the explanatory variable, all variables are logarithmic to simplify coefficients interpretation. The logarithmic form represents the percentage change of the dependent variable as the percentage change of the explanatory variable. The letter l before the variables also indicates the logarithmic variables.
HTE (high-tech export) has been considered as a dependent variable representing the percentage of high-tech export of total manufactured exports for the selected country. HTEit-1 is a high-tech export of last year, as an independent variable appears on the right side of the equation.
NB[8] is the explanatory variable that is used as the national brand value of the selected countries published by the Brand Finance Institute. It should be noted that the this Institute, established in 1996 and headquartered in the UK, is one of the trademark rating institutions in various fields, e.g. auto parts, information technology, civil engineering, food, etc. Each year, having measured the total value of brands available in different countries, the Institute calculates the brands’ intangible asset value, and ranks and publishes them as the National Brand Value. In addition, the published index consists of three main sectors: (i) products and services, (ii) investment, (iii) society, which each of them are subdivided into tourism, market, government, people, and skills, and also the subdivisions are divided into individual criteria. Finally, the score for each criterion is calculated from 100, and by summing the scores of all criteria, the score of the overall national brand index is obtained, which is assigned a score of AAA+ (brand exceptional strength) to d (brand failure). Therefore, based on the mechanism outlined above, with the improvement of the national brand, the high-tech export is expected to be expanded.
Yet, defining an appropriate model that can explain the changing behavior of high-tech export requires considering the other effective factors that are defined as control variables. These variables are selected based on the theoretical export principles and following empirical studies, and the principle of non-autocorrelation between the explanatory variables. However, since the mechanism of each variable’s effect was is explained in the section of theoretical foundations, it is avoided to discuss it again here:
RER stands for the real exchange rate (Rasoulinezhad & Kang, 2016), and is expected to have a positive effect on high-tech export.
IPR is the intellectual property rights following Bayraktutan and Bidirdi (2018) and Kabaklarli et al. (2017) study, in the research model.
OPEN denotes the economy’s degree of openness that follows Mehrara et al. (2017), Sandu and Ciocanel (2014), and Tebaldi (2011) study, in the research model.
FDI represents attraction of foreign direct investment follows the study Garces and Adriatico (2019), Bayraktutan and Bidirdi (2018), Kabaklarli et al. (2017), Gökmen and Turen (2013), and Tebaldi (2011).
In addition, high-tech export data, real exchange rates and the degree of Economy openness are from the World Bank database[9], national brand value data are from the Brand Finance Institute database[10], intellectual property rights[11] data are from the World Intellectual Property Organization database[12] and Foreign direct investment data have been extracted from the United Nations Conference on Trade and Development (UNCTAD) database.
[6]. Eight Islamic member states of the D8 (Indonesia, Iran, Bangladesh, Pakistan, Turkey, Malaysia, Egypt and Nigeria), and six member states of the Persian Gulf Cooperation Council (United Arab Emirates, Bahrain, Saudi Arabia, Oman, Qatar and Kuwait).
[7]. Seven developed economies of the G7 group (Germany, United States of America, United Kingdom, Italy, Japan, France, and Canada), and five emerging economies of the Brix Group (South Africa, Brazil, China, Russia and India).
[9]. http://data.worldbank.org/data-catalog/world-development-indicators
[10]. https://brandfinance.com
[11]. Intellectual Property
[12]. https://www.globalinnovationindex.org