*4.2. The Influence of Directors' Postgraduate Education Characteristics on a Firm's R&D Investment by Industry*

Giroud and Mueller have justified that the product market competition should be considered in order to determine the relationship between corporate governance and R&D policy [46]. Therefore, it is necessary to explore the relationship between directors' postgraduate education characteristics and a firm's R&D investment by individual industry since each industry may face different competitions. Due to the limited number of companies in some industries, this study further recategorized industries into the first industry (11 companies, including agriculture, forestry, animal husbandry, and fishery industry), the second industry (1066 companies, including industrial industry), and the third industry (297 companies, including commercial, financial and real estate, public utilities, culture, education, sports, and IT industry). The first industry was excluded from further analysis due to its small sample size.

The model settings for analyzing the second and the third industry are similar to that shown in Table 5. As shown in Table 6, in the second industry, the higher the percentage of directors with master's degrees and doctoral degrees, the more investment to corporate research and development, controlling directors' work experience in R&D positions, and corporate asset-liability ratio. However, the landscape in the third industry shows substantial distinctions: with the same variables controlled, no significant relationship between board directors' education characteristic and the RD investment of a firm was observed.


**Table 6.** OLS regression results of the influence of directors' postgraduate education characteristics on a firm's R&D investment by industry.

Note: t statistics in parentheses. \* *p* < 0.1, \*\* *p* < 0.05, \*\*\* *p* < 0.01.

#### *4.3. Influence of Directors' Postgraduate Education Characteristics on IT Companies' R&D Investment*

The R&D expenditure in the IT industry is almost two times the average level (seen in Figure 1), which plays a decisive role in the overall level of corporate R&D investment in China. Although the previous analysis on all companies indicates that directors' postgraduate education has no significant impact on R&D investment efforts in a company, it is still necessary to analyze this relationship for IT companies due to the large amount of R&D expenditure and distinctive business feature.

According to the regression results in Table 7, the percentage of directors with master's degrees has a negative correlation with R&D investment in the IT industry. That is, the higher the percentage of directors with master's degrees, the less money a company spends on R&D activities. Additionally, the percentage of doctoral degree holders on board, discipline background, graduation

institutions, and previous work experience in R&D positions has no significant influence on R&D investment expenditure.


**Table 7.** Regression results of the influence of directors' postgraduate education characteristics on R&D investment in the IT industry.

Note: t statistics in parentheses. \* *p* < 0.1, \*\* *p* < 0.05, \*\*\* *p* < 0.01.

#### *4.4. Robustness Test*

Endogeneity is a common plague in research involving corporate governance, firm policy, and performance [47]. There are many unobservable factors that may affect the hiring decision and R&D policy simultaneously, such as board directors' capacity, the firm culture, and so on. Coles and Li [48,49] identified that unobserved attributes of managers have a strong empirical association with riskier corporate policies, such as a higher R&D investment amount. Capacity of board directors can also lead to endogeneity problems. It is common that board directors' capacity is not only associated with their probability of attending postgraduate education, but also has a relationship with the firm's policy in terms of R&D investment. Thus, it will be biased to quantify the relationship between the director's postgraduate education characteristics and the firm's R&D investment without considering these variables. To address the endogeneity problem, Li suggested an extensive range of statistical methods and compared both the advantages and disadvantages of each method [50]. This paper chose to use control variables and the IV method to treat endogeneity issues, considering the nature and the availability of the data.

Corporate asset is directly associated with board members' capacity. Thus, we further added corporate asset into the statistic model as a control variable. Regression results reported in Table 8 suggest our estimation is robust after controlling additional firm attributes.

IV estimation is commonly used to mitigate the endogeneity problem caused by omitted variables. Using the IV approach needs to meet two conditions, i.e., IV must related to the endogenous explanatory variable and unrelated to the disturbing term. Previous engagement in R&D work is highly related with postgraduate education experience. The data from the Chinese Science and Technology Statistics Yearbook shows: 35.62% and 17.73% of people employed by R&D institutions are master's and doctoral degree holders [51]. Nonetheless, previous engagement in R&D work is usually not the requirement of

being a board director. Additionally, we found the proportion of directors with master's degrees is significantly negatively related to a firm's R&D investment in the IT industry, which is different from the rest of the industries. Therefore, we used engagement in R&D work as the IV to test the robustness of our research results in the IT industry. IV 2SLS regression results are presented in Table 9. The first stage regression shows IV is highly related to the endogenous variable (p = 0.00). F statistics (F = 25.82) is greater than the normal threshold of 10, indicating our IV passes the weak IV test and has strong explanatory power to the endogenous variable. The second stage regression shows the coefficient of a directors' master's degree is still significantly negative. Therefore, the robustness of our regression results can be confirmed here.


**Table 8.** OLS regression results before and after adding firm asset.

Note: t statistics in parentheses. \* *p* < 0.1, \*\* *p* < 0.05, \*\*\* *p* < 0.01.


**Table 9.** IV 2SLS regression results in the IT industry.


**Table 9.** *Cont*.

Note: t statistics in parentheses. \* *p* < 0.1, \*\* *p* < 0.05, \*\*\* *p* < 0.01.

#### **5. Discussion and Conclusion**

Based on a sample of 1374 listed companies in China, this study analyzed the influence of directors' postgraduate education characteristics on R&D investment.

First, the R&D investment of China's listed companies needs further improvement and its distribution among different industries is highly uneven. The average proportion of a Chinese firm's R&D expenditure was 5.10% in 2017, which indicates a considerable R&D gap between China and the United States since the average R&D spending of American firms reached 14.5% from 1995 to 2015 [52]. Enterprise is the main driving force of a national innovation system. With the intention of building an innovative country, corporations should be encouraged to increase their R&D investment. When they are broken down to different industries, listed companies could be classified into four tiers, where IT industry was in the first tier, with R&D investment exceeding 10%. The second tier covered industrial (4.74\$) and culture, education, and sports (4.32%). The next tier included agriculture, forestry, animal husbandry, and fishery (2.73%), public utilities (1.88%), and financial and real estate industry (1.69%). Commercial industry was classified into the fourth tier since its R&D expenditure was less than 1% (0.92%).

Second, the proportion of doctorates contributes significantly to the amount of R&D investment in firms, with the corporate asset-liability ratio, industry, and director's working experience being controlled. This finding helps to confirm the hypothesis that highly-educated directors are likely to increase the firm's R&D investment. As Wally and Baum noted, higher levels of education usually means greater cognitive ability [53], which allows for higher ambiguity toleration [36], better understanding of new ideas [37], more creative solutions to tackle complicated situations [54], all of which pave the way to increasing R&D investment. In respect to discipline background, science and engineering education experience may equip directors with innovation management knowledge and skills so that they gain more confidence to invest more on R&D activities [55]. But the graduation institution does not seem to have impact on directors' decision towards R&D investment. On one hand, this finding confirmed the important role of doctoral education in promoting a firm's R&D investment. On the other hand, the 'Top End Effect' shall not be excessively emphasized, and overseas returners may not bring in more R&D investment. The influence from the non-elite postgraduate education institutions are essential to the improvement of a firm's R&D investment as well.

Third, the influence of directors' postgraduate education characteristics on their firm's R&D investment generally occurs in the second industry. The higher the percentage of directors with postgraduate education degrees is, the more R&D investment will be. This is consistent with previous research conclusions based on the data from China's manufacturing industry [56]. Compared with other industries, the percentage of directors with master's degrees in industrial companies is relatively lower and has space for improvement. Therefore, enhancing the postgraduate education level for

directors in industrial enterprises, especially increasing the percentage of directors with master's degrees on the board, may be one of the strategies to enlarge R&D investment for this industry.

Fourth, the high proportion of directors with master's degrees in the IT industry prevents corporate R&D investment. After employing the IV approach, the result is still robust. To find out the possible explanation, we checked the original data of the IT company with the lowest R&D investment (code = 300226). Results show it only invested 0.06% of total sales in R&D activities and its asset-liability ratio was up to 78%. As for its board configuration, all the directors are master's degree holders and majored in economics and management. This typical case indicates some IT companies tend to develop through raising capital instead of R&D innovation. Descriptive statistics reveal that nearly half (48%) of directors in this industry hold master's degrees, whereas the percentage of doctorate holders is only 24 percent. In terms of discipline background, 42% of directors majored in economics and management and only 20% majored in science and engineering. This may explain the negative coefficient between the percentage of master's degree holders and R&D investment in the IT industry.

#### **6. Implication and Limitation**

To summarize, this paper contributes to related research both theoretically and practically. In term of theoretical contribution, our findings extend the streams of literature concerning innovation management theory, upper echelon theory, and human capital theory by examining how the firm's R&D investment is affected by postgraduate education experience of directors, such as their education level, discipline background, and graduation institution. In contrast to prior innovation management research highlighting external determinants of corporate R&D innovation, our research finds the higher the percentage of directors with doctoral degrees, the more investment in R&D activities. This provides empirical evidence that internal factors from the board of directors also play an important role in R&D activities. In line with the research from the perspective upper echelon theory, this paper pays attention to a director's individual characteristics as well, especially educational characteristics. Our research findings indicate the association between a director's postgraduate education characteristics and a firm's R&D investment varies in different industry environments. In the second industry, which mainly consists of industrial companies, the higher proportion of postgraduate education degree holders as directors, the more expenditure in R&D activities. However, this correlation is not statistically significant in the third industry, which involves commercial, financial and real estate, public utilities, culture, education, sports, and IT companies. Through the lens of the human capital theory, there is economic value embedded in postgraduate education, which is justified in this paper in respect to R&D investment. In addition to the positive impact of human capital on economic growth and R&D innovation that has been studied thoroughly, our contribution lies in unpacking the complexity in reality due to certain attributes of both board directors and corporate developing strategies. Findings from IT companies confirm that, if a capital-driven strategy is adopted, directors with master's degrees and an economics and management background tend to reduce R&D investment. In term of practical contribution, our research findings lead to several prominent implications for enhancing the whole R&D investment of China's listed companies. Enterprises should realize the correlations between directors' postgraduate education characteristics and R&D investment, and then adopt targeting innovative strategies through board staffing and configuring. For the achievement of sustainable innovation of a state, national policy should also encourage universities to increase the availability of innovation and entrepreneurship education in the curricula system of postgraduate education.

Limitations of this paper and further research directions mainly lie in the following two aspects. First, R&D activities are highly risky and demand long-term investment [57]. Due to the fact that a firm's R&D investment has path-dependent inertia to some extent, tracking and analyzing synchronic rules of an enterprise's R&D investment will be the next research emphases. What's more, other factors like firm culture and technical advancement may also affect R&D investment and board hiring decisions. Limited by the availability of data in the CSMAR database, these factors are not considered in this

paper. Future research may collect additional companies, and also extra variables to better model their contributions to R&D investment.

**Author Contributions:** Conceptualization, C.W. and C.N.; methodology, C.W., J.Y. and Z.C.; writing—original draft preparation, C.W. and J.Y.; writing—review and editing, C.W., J.Y., Z.C. and C.N.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.
