Keywords

7.1 Introduction

The classical growth model developed by Solow (1956) predicts that GDP per capita would converge in the long run. As Hall and Jones (1999) pointed out, however, the income gap between rich and poor countries remains wide. Jones and Romer (2009) in their paper titled “New Kaldor Facts: Ideas, Population and Human Capital” state that the income gap can be attributed to differences in productivity as one of the stylized facts. Productivity differences can be considered not only at the country, but also at the firm level. Empirical studies since the 1990s also found that the productivity gap continues to exist at the firm and establishment levels. Bartelsman et al. (2009, 2013), in constructing the productivity database at the firm level across countries, investigate the determinants of these productivity gaps and find that activating entry and exit of firms and more efficient resource reallocation would improve the economy-wide productivity.

The economic performances in Japan and Korea, which both suffered from the financial crises in 1997 and successive deep recessions are also different. While the Japanese economy has stagnated for two decades due to large non-performing loans, the Korean economy recovered rapidly. As a result, firm performance in Korea caught up with those of Japan in some competing industries such as electric machineries and electric devices as shown in Fukao et al. (2008). In growth accounting using the framework of McGrattan and Prescott (2005, 2010), Miyagawa and Takizawa (2011) showed that the labor productivity gap between Japan and Korea after the financial crises was explained by the difference in TFP growth.

Many studies at the firm level found that Korean firms are rapidly catching up with Japanese firms in terms of productivity and market shares in several sectors. Jung et al. (2008) noted that while the productivity of Korean firms were as low as half of that of Japanese firms in the mid 1980s, and there had been substantial catch-up in productivity by the listed Korean firms which were now on average within the 10 % range in the late 1990s. Jung and Lee (2010) find both sectoral-level and firm-level factors responsible for the productivity convergence. While explicit knowledge oriented sectors, like IT, tend to show faster catch-up, firm-level factors, such as innovation capability and export-orientation, were also significant. Joo and Lee (2010) compare Samsung and Sony in terms of the various indicators created using patent data including citations, and conclude that while Samsung caught up with Sony in the mid 2000s in terms of market capitalization and sales volume, technological catch-up, in terms of patent count, quality and mutual citations, etc, occurred as early as the mid 1990s.

While the reasons behind Korea catching up to Japan should involve many diverse factors, the existing studies tend to consider mostly tangible aspects of the firms which are often reflected in standard financial statements or patent application data. In the survey of research on micro-level productivity, Syverson (2010) divides factors that affect changes in productivity into intra-firm factors and environmental factors external to the firm. He argues that one of the main intra-firm factors is in management practices. Aoki (2010) also emphasizes that the organizational architecture within a firm is a major driver of the corporation system in each country. These arguments are in line with the seminal paper by Bloom and Van Reenen (2007) in which they conducted their own survey on management practices of four advanced countries (France, Germany, the UK, and the US) and examined the relationship between the management score and firm level productivity. In their paper, management practices were converted to scores based on interview results, and these scores were included as independent variables when they estimated the production function. The key finding in their paper is that there is a significant difference in management score among countries surveyed. US firms got the highest score of the four countries. They believed that the low score in continental European firms was partly explained by weak competition and the prevalence of many family-owned firms. The estimation results showed that the productivity differences corresponded to the differences in average management scores.Footnote 1 In Japan, Kurokawa and Minetaki (2006), Kanamori and Motohashi (2006), and Shinozaki (2007) examined the effects of organizational reform resulting from IT investment on firm performance by using the Basic Survey on Business Enterprise Activities and the IT Workplace Survey. Their studies suggested that organizational reform resulting from IT investment was partially responsible for improving firm performance. This study extend the previous studies by focusing on comparison of management practices in Korean and Japanese firms in a same framework.

The purpose of this paper is to compare the management quality between Korea and Japan that is considered as a cause for productivity differences in the two countries. We conduct a similar interview survey with respect to Korean and Japanese firms to Bloom and Van Reenen’s survey, and compare the organizational and human resource management practices in firms of the two countries. In the next section, we describe an outline of our interview survey. In the third section, we construct a management score by quantifying the two interview results of Japanese and Korean firms, and compare the management practices. Although our interview survey basically follows Bloom and Van Reenen (2007), we incorporate some questions that were not included in their surveys to capture some unique features of Japanese and Korean firms such as the role of informal meetings within the firm and the speed of decision-making. Hence, we compare not only management scores that represent organizational and human resource management practices but also other management styles between Korea and Japan. In the last section, we summarize our studies and discuss the future studies on management practices and firm performances.

7.2 Outline of Interview Surveys in Japan and Korea

The surveys on management practices in Japan and Korea were conducted twice: Once in 2008 and the second time in 2011–2012. The description of each survey is summarized in Table 7.1. The two Japanese surveys were conducted by Research Institute of Economy, Trade & Industry (RIETI) The first Korean survey was conducted by Japan Center for Economic Research (JCER), and the second Korean survey was conducted by Samsung Economic Research Institute. The second Japanese survey was originally scheduled to be conducted in 2011, but was postponed to 2012 due to the Great East Japan Earthquake in March 2011. The number of responses in the second survey was drastically lower than the first, because of the earthquake and the fact that the interviewees were limited to publically traded firms.

Table 7.1 Outline of surveys

In our study, we followed the interview survey conducted by Bloom and Van Reenen (2007). However, we conducted the survey by meeting the managers of the planning departments of firms face-to-face, while Bloom and Van Reenen (2007) conducted their survey by telephone. The reason that we conducted face-to-face interviews is that we were afraid of low response rates. In Japan and Korea, when we want to ascertain qualitative features in firms, face-to-face communication is a more useful tool than telephone interviews.

Bloom and Van Reenen (2007) classified their 18 interview questions into four categories: product management, monitoring, the firm’s target, and incentives for workers. While their survey was extended to only manufacturing plants, our survey was also extended to firms in the service sector. Thus, we excluded questions about product management in the service sector. As a result, we can classify our questions into two categories: organizational management and human resource management. In the first category, we wanted to examine the organizational goals, communication within the firm, and organizational reform. The second category about human resource management covers questions on promotion and training programs.

The interview also includes questions that are not directly related to management practice and human resource management. Since the IT revolution, changing a pyramid-type decision-making process into a more flat process became more popular. We ask questions targeted to help our understanding of whether firms underwent such organizational restructuring that includes the decision-making process. In the first round, we also ask about the vision of the firm. In the second round of interviews, considering the globalization that was taking place, we include questions regarding firms’ primary market and competitiveness (the number of competitors), and the time it takes to enter and exit businesses. The detailed interview questions are shown in Appendixes 1 and 2.

For each question, we have three sub questions. The structure of the point system is that the more sub-questions answered positively in each large question, the more points you score, for instance, in human resource management. In each question with three sub-questions, you score 4 points if you answer positively to all of the three sub-questions. Similarly, with positive answers to the first two sub-questions only, you would score 2 points. In other words, we quantify the responses to the above questions as follows: If the firm manager responds negatively to the first sub-question, we give the response 1 out of a possible total of 4 points for the entire question and move to the next question. If he responds positively to the first sub-question, we move to the second sub-question. If the manager responds negatively to the second sub-question, we mark a 2 and move to the next question. If he responds positively to the second sub-question, we move to the last and third sub-question. In the last sub-question, if the manager responds with a positive answer, he is given 4 points for the positive responses for all three sub-questions while a negative response is given a point of 3 for the two previous sub-questions he answered positively.

7.3 Comparison of Results in Interview Surveys Between Korea and Japan

7.3.1 Distribution of Sample Firms by Sector and by Size

We first show the distribution of the firms interviewed. Table 7.2 provides the share of firms in the manufacturing sector and the service sector for both surveys. In the first survey, we interviewed machinery industries for the manufacturing sector and information service, media-related industry, and retail industry for the service sector. In the second survey, since the sample was limited to public traded firms, we did not limit the interview to specific industries. As the first Japanese survey focused on specific industries in the manufacturing sector, the share of firms in the manufacturing sector is relatively small with the ratio of the manufacturing to service sector being 1–2. In the second survey, the ratio is reversed. In both of the Korean surveys, on the other hand, the manufacturing sector constitutes approximately 80 % of the interviewees.

Table 7.2 Distribution of industries surveyed

Table 7.3 shows the distribution of firms by size. In the Japanese surveys, small and medium sized firms with less than 250 employees constituted a slightly greater than one-third share of the sample. On the other hand, in the first Korean survey, the share of small and medium-sized firms dominated the survey and accounted for 65 % of the sample. In the second Korean survey, however, this share fell to 46.3 %.

Table 7.3 Size distribution of surveyed firms

7.3.2 Comparison of Management Scores

We now compare management scores. In the first survey, we take the average of these scores in Q2, Q4, Q5, and Q7-13 to obtain an overall management score. In the second survey, we take the average of the scores assigned to Q3, Q4, Q5, Q6-1, Q10-2, Q10-3, and Q10-4 to obtain the overall management score. The organizational management scores are the average scores in Q2 in the first survey and by the average score in Q3 and Q10-1 in the second survey. Lastly, the human resource management score is the average score in the questions that are not related to organizational management. A high management score implies that management targets within a firm are set and are widely recognized by the employees. On the other hand, the human resource management score is high when employees with high performance receive rewards and promotion swiftly, and when firms invest in human capital accumulation.

Table 7.4 shows the management scores in Japan and Korea. In both surveys, management scores in Japanese firms are higher than those in Korean firms except for the case of management scores in large firms in the second survey. However, the Japanese management score falls slightly between the first and second survey while the Korean management score increases greatly between the first and second survey, catching up with Japan. In particular, the management score in Korean large firms surpasses that in Japanese large firms in Japan.

Table 7.4 Management scores based on the interview surveys

Figures 7.1, 7.2, 7.3, and 7.4 show the distribution of scores in all firms and all interview questions in Japan and Korea by using Kernel density. We find that the distributions of management scores in Korean firms are more dispersed than those in Japanese firms. This implies that there are many high score firms and low score firms in Korea while management scores in Japanese firms are more concentrate around their mean values. When we compare the distributions in the two surveys in both countries, the distributions in the second survey in Japan shift slightly to the left. On the other hand, the distributions in the second survey in Japan shift drastically to the right.

Fig. 7.1
figure 1

Distribution of management scores (all firms in the 1st survey)

Fig. 7.2
figure 2

Distribution of management scores (all firms in the 2nd survey)

Fig. 7.3
figure 3

Distribution of management scores (1st vs. 2nd in Japan)

Fig. 7.4
figure 4

Distribution of management scores (1st vs. 2nd in Korea)

We check similar distributions by type of management and by firm size (see Figs. 7.5, 7.6, 7.7, 7.8, 7.9, 7.10, 7.11, and 7.12). The distributions of organizational management scores and human resource management scores show similar patterns to Figs. 7.3 and 7.4. When we compare the two surveys, the distributions of two types of management scores in the second survey in Japan do not change much. However, the two distributions of the management scores in the second survey in Korea shifts greatly to the right. In the case of small and medium sized firms, we also see similar patterns for other cases.

Fig. 7.5
figure 5

Distribution of organizational management scores (1st vs. 2nd in Japan)

Fig. 7.6
figure 6

Distribution of organizational management scores (1st vs. 2nd in Korea)

Fig. 7.7
figure 7

Distribution of human resource management scores (1st vs. 2nd in Japan)

Fig. 7.8
figure 8

Distribution of human resource management scores (1st vs. 2nd in Korea)

Fig. 7.9
figure 9

Distribution of management scores in large firms (1st vs. 2nd in Japan)

Fig. 7.10
figure 10

Distribution of management scores in large firms (1st vs. 2nd in Korea)

Fig. 7.11
figure 11

Distribution of management scores in SME (1st vs. 2nd in Japan)

Fig. 7.12
figure 12

Distribution of management scores in SME (1st vs. 2nd in Korea)

The distributions of management scores by firm size show similar patterns to those by type of management. In large firms, the distribution of management score in the second survey in Korea showed a great shift to the right, while that in Japan shifted slightly to the left. These shifts suggest that the means of the management scores in Korean large firms is greater than that in Japanese large firms. In addition, these results support the notion that the performances in the listed Korean firms surpassed those in the Japanese listed firms, as Jung et al. (2008) and Joo and Lee (2010) showed. Similarly, the distribution of management scores of the Korean SMEs shifts to the right, although the distribution of management scores in the Japanese SME in the second survey does not move much.

We check the two distributions between the first and second surveys of Japan and Korea by using the Kolmogorov-Smirnov test. Suppose the two cumulative distribution functions (F(x) and G(x)) and take the maximum differences between two distributions (D mn ) defined from the sample distribution functions of F(x) and G(x).

$$ {D}_{mn}= \sup {}_{-\infty <x<\infty}\left|{F}_m(x)-\left.{G}_n(x)\right|\right. $$

In the Kolmogorov-Smirnov test, the null hypothesis is that the two distributions are the same (F(x) = G(x)). If the test statistics \( {\left(\frac{mn}{m+n}\right)}^{1/2}{D}_{mn}>c \) and c is an appropriate constant, the null hypothesis is rejected.

The test results are shown in Table 7.5. The Kolmogorov-Smirnov test is conducted in four cases: the comparison of two distributions in Japan and Korea in the first survey, the comparison of two distributions in Japan and Korea in the second survey, the comparison of two distributions in the first and second surveys in Japan, and the comparison of two distributions in the first and second surveys in Korea. In the first row of the table, we test the hypothesis of whether the sample values in Japan are significantly smaller than those in Korea. ‘Distance’ in the second column shows the maximum distance in the case where the sample value in Japan is less than that in Korea. P values in the first and second surveys show that sample values in Japan are not significantly smaller than those in Korea. However, the P value in the third table shows that the sample values in the first survey are significantly smaller than those in the second survey in Japan. In the case of Korea as well, the sample values in the first survey are significantly smaller than those in the second survey. The second row of the table tests the opposite case. The Kolmogorov-Smirnov test shows that sample values in Japan are significantly larger than those in Korea in both surveys. In the case of the first and second surveys in Japan, the sample values in the first survey are significantly larger than those in the second survey in Japan. The last row shows the combined results of the previous two tests.

Table 7.5 Kolmogorov-Smirnov test

These tests imply that the distributions of management scores in Japan have shifted significantly to the right more than those in Korea in both interview surveys. Overall, the organizational targets are clear to all employees in Japan in more cases than in Korea, or Japanese firms improve their organizational structures more aggressively than Korean firms, because high scores in organizational management indicate a greater degree of transparency of organizational goals or aggressive organizational reform. As for human resource management, Japanese firms are more flexible in their human resource management than Korean firms. However, the flexibility of human resource management improves in Korea, while it declines in Japan.

However, the difference in management scores partially reflects the difference in samples between the first and the second survey. When we limit the sample to be the same and consistent between two surveys for Japan, the management scores in consistent samples show similar patterns to those in the entire sample in the second survey (see Table 7.6). This implies that the shrinking gap of management scores between Japan and Korea cannot be entirely attributed to changes in the samples.

Table 7.6 Comparison of management scores in corresponding industries in the first and second surveys in Japan

7.3.3 Features of the Results in the Second Survey

In the second survey, we asked additional questions to shed some light on management styles in Japanese and Korean firms and the market conditions that they face. Table 7.7 summarizes the responses to the supplementary questions. Question 1 in the second survey asks outline of business and several economic environments which Korean and Japanese firms face. As for the first question in Question 1 which asks the main market for a firm, almost two thirds (2/3) of the Japanese firms answered that they sell more than 75 % of their products in their domestic market. On the other hand, less than half of the Korean firms are selling more than 75 % of their products in their domestic market (see Table 7.7a). The competitive environment asked in the fourth question in Question 1 also differs greatly between Japan and Korea. While more than half of the Japanese firms have more than six competitors, approximately half of the Korean firms have fewer than five competitors (see Table 7.7b).

Table 7.7 Summary of responses to additional questions in the second interview survey

It is often argued that the major difference between Japanese and Korean firms is the speed of the decision-making process. In the second survey, we ask questions related to this issue. Table 7.7c illustrates that time it takes much less time to change the existing targets of the firm for Korean firms than for Japanese firms based on the second question in Question 3-3-4. While 65 % of the Korean firms revise organizational goals and other production processes within 3 months, it takes more than 6 months for approximately three fourths (3/4) of the Japanese firms to reach similar decisions.

However, in Table 7.7d constructed from Question 9-1, over 40 % of the Japanese firms responded that less than 20 % of the time before the project begins is spent on the “nemawashi” informal consensus building. On the other hand, approximately 60 % of the Korean firms spend 20–60 % of their time on this consensus building. Yet, considering the time it takes for the Japanese firms to change its targets, we cannot conclude that the time for these preparations is shorter for Japanese firms than Korean firms. Based on the information in Table 7.7c, suppose that the time it takes for the organizational decisions to take place is 10 months for Japanese firms. The time spent on the informal consulting building constitutes 20 % of that time, which is 2 months. On the other hand, the Korean firms spend on average 3 months to reach organizational decision. Then, 40 % of the 3 months would be spent on this consensus building, which would be 1.2 months.

However, Table 7.7e constructed from the results of the supplementary question in Question 9-1 shows that the time it takes to decide on new projects is not necessarily shorter for the Korean firms. In Japan, the cases can be extreme: cases in which decisions would be made very quickly and those in which decisions would take more than 6 months. On the other hand, almost 60 % of Korean firms take more than 6 months to make a decision on new projects. A similar trend is observed when it comes to the termination of existing projects in Table 7.7f constructed from the results of the supplementary question in Question 9-2. Contrary to common perceptions, a large fraction of Japanese firms take less time to make decisions than their Korean counterparts.

Lastly, we compare the amount of information shared with the manager at establishment level by using the results in Questions 9-3 and 9-4. Table 7.7g indicates that only 12 % of the Korean firms share less than 40 % of pertinent information to project managers while 40 % of Japanese firms do. This indicates that overall, Korean firms tend to give more decision-making authority to project managers than Japanese firms and that decentralization is more common in Korea. Table 7.7h also shows that the share of information obtained by a project manager is via informal route and is not necessarily higher in Japan than in Korea.

7.4 Conclusions and Discussions

In the last 20 years, Korean firms have been catching-up with Japanese firms and some firms have already surpassed the Japanese firms in productivity or market shares. According to growth accounting in Japan and Korea, the accumulation of intangible assets has played a key role in explaining the difference in economic performance in the two countries. Among several types of intangibles, management skills and human capital are crucial to the improvement of a firm’s performance. Bloom and Van Reenen (2007) examined the effects of organizational and human resource management on firm performance using interview surveys conducted in France, Germany, the UK, and the US. Following their study, we conducted interview surveys on organizational and human resource management in Japan and Korea.

In this paper, we constructed scores on management practices in each firm based on the two interview surveys. For the scores in organizational management, firms that have clear organizational targets, better communication amongst employees, and conduct organizational reforms would have a higher score. For the scores in human resource management, firms that evaluate human resources flexibly and strive to keep employees motivated would mark higher scores.

When we compared the distributions in average management scores between Japanese and Korean firms, the mean value in Japan was higher than that in Korea except in the case of large firms in the second survey. However, the gap in management scores between Korea and Japan shrunk significantly in the second survey, because management practices in Korean firms improved compared to their scores in the first survey to the second survey. In particular, the higher mean value of management scores seen in large firms in Korea than those in Japan in the second survey is consistent with recent studies comparing firm performances in large firms in Korea and Japan.

Kolmogorov-Smirnov statistics comparing the distributions between Korea and Japan show that the distributions in the average score in Japan are significantly different from that in Korea. The results show that the overall management quality in Japanese firms is higher than that in Korean firms. However, the distribution of management scores in Korean firms has shifted to the right, while that in the Japanese firms moved slightly to the left. In particular, the distribution of the human resource management score of Korean firms shifted to the right in the second survey, which suggests that there was the improvement in human resource management in Korean firms.

Our next task is to examine whether the improvement in firm performance is associated with better management practices. Lee et al. (2012) have already examined the relationship between management practices and productivity at the firm level by conducting production function regressions using the results in the first interview surveys. They show that management scores are positively associated with productivity. In particular, better human resource management is positively correlated to the productivity. From these results, we expect that improvement in human resource management in Korean firms in the second survey is associated with productivity improvement in those firms and the slight decline in management scores in the second survey in the Japanese firms is related to stagnating productivity in those firms. After constructing a database for the management scores in the second survey and corresponding financial statements in the sample firms, we will be able to conduct more rigorous and comprehensive verification of the relationship between the management practices and productivity performance.