Chapter 1 Introduction

Chapter 2 Preliminary Research and Data

Chapter 3 Analysis of Productivity and Employment Effect of R&D Activities

Chapter 4 Analysis of Productivity and Employment Effect of National R&D Projects

Chapter 5 Analysis of Patents Using Natural Language Processing

Chapter 6 Conclusion and Policy Implications

This report aims to analyze the economic effects and labor market outcomes of patents using two approaches: traditional econometric methods and natural language processing. Patents are one of the outcomes of research and development (R&D) activities, serving as an indirect indicator of R&D or productivity.

In this study, rather than measuring the performance or investment level of R&D activities by patents, we used the R&D investment costs of companies and the “R&D Activity Index” (R&D AI) developed by Korea Rating & Data (KoDATA). The analysis showed that R&D activities measured in terms of R&D investment costs exhibited a positive correlation with productivity measured by per capita sales. However, the R&D AI has a negative correlation. These findings could stem from differences in measurement methods despite the high correlation between the two indicators. Additionally, the R&D AI reflects the long-term sustainability of R&D activities, while R&D investment costs represent short-term expenditures. The increase in R&D expenditure with improved financial performance may be perceived as a direct correlation between R&D investment costs and productivity.

However, an employment indicator measured by the number of employees showed a positive correlation with both the R&D AI and R&D investment costs. This suggests that because R&D professionals have different educational and skill requirements compared to other types of white-collar workers, increasing new hires may be necessary to expand R&D activities. There may be a reverse causality in which the need to expand business and maintain market share drives the expansion of R&D activities. Which of these pathways is greater remains to be fully elucidated.

Moreover, both the R&D Activity Index and R&D investment costs, below a certain threshold, evince negligible associations with indicators such as productivity, sales, or employment. This suggests that R&D activities need to reach at a certain scale or higher in order to yield outcomes. (For the full text of the English summary, please download the “Download 2 English_Abstract”.)