Research on the Impact of Global Production Segmentation on the Income of Economic Competencies

— This paper investigates the role of global production segmentation for the income of economic competencies (ECs) such as design, brand name, organizational capital at the country level for 29 countries from World Input-Output Database (WIOD) over 2000-2014. This study distinguishes various forms of production segmentation and examines how various integrated forms co ntribute to countries’ gain of ECs.This paper finds that production lengthbased on forward industrial linkage (plvg) is positively correlated with value appropriation ofECs in global value chain (GVC), while production length based on backward industrial linkage (plyg)exhibits an opposite effect on the value creation of ECs. Based on whether production happens in domestic factories or foreign partners, this paper finds that the positive effect of plvg on ECs originates from the extension of international production portion, while the domestic production dominates the negative impact of plyg on ECs. rk LM statistics and Kleibergen -Paap rk F statistic below the square brackets are the P value and the critical value at the 10% statistical significance level. The results show that the instrumental variables used in this paper are acceptable.

One of the main purposes of this paper is to analyze the changes in the income distribution of ECs, so that we can intuitively judge whether the competitiveness of developing countries in terms of brand, organizational assets and design has achieved convergence to developed countries. Secondly is to analyze the role of GVC in the global redistribution of income of ECs, and to explore the impact of different value chain integration forms on ECs.
The rest of this study is structured as follows. Section 2 introduces the measurement method of ECs and data sources and provide some descriptive evidence on the redistribution of the return of ECs. Section 3describes empirical strategy; Section 4presents econometric results and discussion on endogeneity. Section 5 provides some robust tests and Section 6 concludes.

Calculatingthe return of economic competencies
The estimation method of ECs can be found in Karabarbounis and Neiman(2018)and is presents as equation (1): In equation (1) Barkai (2020) and is as equation (2): In equation (2), represents depreciation rate, denotes expected inflation rate, and is the nominal interest rate.

III. EMPIRICAL STRATEGY
Based on the above analysis, this research constructs the following econometric model to examine the impact of production segmentation on the return of ECs: In equation (3), the subscripts and year have the same meaning as equation (1); represents the logarithm of the ECs income at the national level.
is the production length based on the forward industrial linkage; denotes production length based on backward industrial linkage. X is the set of control variables of country's specific characteristics that may affect the income of ECs; and represent ② RUS ECs income share also decrease in 2014 compared to the corresponding figure in 2004 which is the earliest data we gained in this paper.
individual and time fixed effects, respectively; is the random error term. All the production length variables are in logarithmic form.

Analysis of benchmark regression
Thepanel data is unbalanced. After Huasman's test and auxiliary regression estimation for equation (1) (1) resulting in measurement problem in the regression results.
The instrumental variables and robustness tests will be used to further confirm whether the effect exist or not. In Column (3) in Table 1  Generally, the use of instrumental variables does not affect the interpretation of the results in benchmark regression.

V. ROBUSTNESS CHECK
The estimation of ECs income is affected by the choice of rental rate of capital and depreciation rate.
Excessive depreciation rate will lead to overestimation of capital share and underestimation of ECs income, while underestimation of real interest rates will be the opposite.
Since the PWT database already provides country-specific depreciation rates, this paper uses the short-term lending interest rate data from FRED and exploiting higher actual interest rates to re-estimate the income of ECs in order to  Higher capital rental rates can be seen as compensation for risk premiums and asset liquidity, which can further alleviate the concern questioned by Karabarbounis and Neiman (2018) that using risk-free interest rates to calculate the cost of capital may lead to an overestimation of ECs income. If there are significant regression differences between the indicators of GVC on the income effects of ECs calculated by different methods, then the previous analysis conclusions cannot be considered to be robust.
Using the short-term nominal lending interest rate (lneclr) and 3% real interest rate (lnecd) to replace the 10-year treasury bond yield rate to calculate the ECsand re-estimate the equation (1), the regression results are reported in Table 3. The regression results show that, except for coefficients of in columns (2) and (5) of