Abstract:     It is a fundamental question in the strategy research of what determines the scope of the firm. Theorists suggest that it is product diversification or geographic diversification which determines the scope of the firm, and appeal that empirical studies should examine both effects in a context outside America and the developed countries. In this study we respond to these calls and examine the effects of different types of firms′ core competences and their influences on the product diversification or geographic diversification simultaneously. With the integration of institution-based view and resource-based view (RBV), we extend the research of this fundamental question to the context of the fast development and distinct path selections of the Chinese private enterprises. Private enterprise is one of the most important engines of economic growth in China since the reform and opening up of China, but we know very little about the process of diversification, one of the most popular development strategy for the Chinese private enterprises, especially the relationship among different core competences and dimensions of diversification strategy which determine the scope of the firm. Because of the legitimacy and liability of newness, most private enterprises in China have the characteristics of new starting firms. Usually, Chinese private enterprises discover and exploit business opportunities by developing capabilities of R&D or political connections, choose different development paths and lead to different scopes of firm in the end, i.e. different level of product diversification or geographic diversification. These processes are influenced by the institutional differences especially different levels of the market liberalization come from the obvious regional segmentation in China. In this study, we select a specific sample from 1291 private enterprises public listed in the Chinese stock market. We delete banks and the other financial companies which operated very differently with most private companies and firms have no obvious development strategy such as special treated firms (i.e. ST and *ST firms) with very poor performance. In the end, we select 1161 private firms in the 19 industries. We measure the product diversification with an entropy index of prime operating revenue and geographic diversification with the Herfindahl-Hirschman index according to the authority literatures. Since product diversification and geographic diversification are continuous variables, we employ multiple linear regression for the analysis using the Stata software, and apply hierarchical moderated regression approach to test all the hypotheses. Hierarchical analysis allows for a comparison between alternative models with and without interaction terms, where an interaction effect only exists if the interaction term contributes significantly to the variance explained in the dependent variable over the main effects of the independent variables. We also check the robustness by different measures of the major study variables, and construct models to examine the endogeneity using the Granger causality test. Our results are robust when examined with different measures of the major study variables and pass through the causality test desirably, it is the capability of R&D or political connections which lead to high-level or low-level product diversification and geographic diversification and not the vice versa. We select 2007 as the data starting year, because in 2007 all listed companies are required to implement the new accounting standards by Chinese government and disclose company information with even more high quality. Empirical results of the private enterprises in 2007-2014 reveal that both core competence, i.e. capabilities of R&D or political connections, have direct effects on product diversification and geographic diversification. Capabilities of R&D lead to high level geographic diversification and more focused product diversification, and political connections lost some significance which imply its mechanism is more complex than the R&D capabilities. However, political connections have significant negative moderate effects which changes the above relationships to the contrary directions. In the firms with high level capabilities of political connections, R&D capabilities make firms more localized and operated in many industries, so they have high level product diversification and low-level geographic diversification. On the contrary, in the firms with low-level political connections, R&D capabilities make firms devoted to the focused industries and enter different regions to make use of the R&D advantages, so they have low-level product diversification and high-level geographic diversification. But with the improvement of the regional institutions measured by the market liberalization, negative effects of political connections are suppressed and R&D effects are protected. In the regions with high-level market liberalization, market mechanism surpasses the effects of the political connections, direct effect of R&D capabilities appears again, private firms select the path of low-level product diversification and high-level geographic diversification. In practice, both capabilities are useful and to fully understand their effects need to consider the special contexts and the strategic outcomes. In the regions where institutions of the market liberalization are not formed it is better to develop capabilities of political connections and build up a local "business empire" which operates in many industries or many product units, i.e. high-level product diversification and low-level geographic diversification. Many firms set up in the county or the 4th- and 5th-tierd cities have the characteristics which deeply connected to the government officials and operates in many industries such as real estate, department store, hotel and catering industry besides their major field. On the contrary, In the regions with more mature of market economy institution, private firms usually devote to R&D capabilities and focus on narrow product field. In this context, political connections are very expensive or not very useful so they are replaced by market mechanism, market efficiency and protection of law help firms to select the path of internal development. In the long run, it is much better to develop capabilities of R&D and push the company to be a specialized competitor operating in the countrywide or worldwide which operates only in several industries or focused product fields, i.e. low-level product diversification and high-level geographic diversification. 

Key words:  R&D capability, political connection, institutional environment, private enterprise, diversification strategy