The Contact-Hitting R&D Strategy of Family Firms in the Japanese Pharmaceutical Industry

Abstract

This article addresses the following question: How do family firms investing less in research and development (R&D) than nonfamily firms compete in R&D intensive industries? Using Japanese pharmaceutical industry data, we found that family firms produce more patents per R&D than nonfamily firms but are not biased toward low-value innovations. Further analyses of the distribution over innovation value suggested that family firms adopt a “contact-hitting R&D strategy,” avoiding radical innovations and pursuing incremental innovations compatible with their signature moves: innovation through tradition and narrow and internal search and resulting in may low-value innovations and a few mid or high-value innovations.

Keywords

innovation productivity signature moves contact-hitting R&D strategy frequency distribution over innovation value

How do family firms, which most empirical studies (e.g., Block, 2012; Block, Miller, Jaskiewicz, & Spiegel, 2011; Chen & Hsu, 2009; Chrisman & Patel, 2012; Munari, Oriani, & Sobrero, 2010) have found invest less in research and development (R&D) than their nonfamily counterparts, compete in R&D-intensive industries? This study attempts to tackle this heretofore insufficiently addressed question by focusing on a novel dimension of innovation and R&D among family firms. This question is highly important because family firms, known to be globally prevalent and performing well (R. C. Anderson & Reeb, 2003; Miller & Le Breton-Miller, 2005; Miller, Le Breton-Miller, Lester, & Cannella, 2007; Porta, Lopez-De-Silanes, & Shleifer, 1999; Villalonga & Amit, 2006), should be playing a crucial role in the R&D and technological innovation that accelerate economic development and firm growth (Schumpeter, 1926; Solow, 1957). As some scholars suggested, one possible answer to this question is that family firms may use R&D resources efficiently (Chen & Hsu, 2009), resulting in a greater level of innovation with fewer R&D resources (Duran, Kammerlander, Van Essen, & Zellweger, 2016; Kammerlander & Van Essen, 2017). If family firms show higher innovation productivity than nonfamily firms, the former can compete in R&D-intensive industries. However, there is not a consensus on why family firms show higher innovation productivity. Therefore, this study attempts to distinguish among the reasons for the high innovation productivity of family firms.

Previous studies have produced different results on innovation productivity and provided various explanations for these differing results. Some scholars of family business argue that because of their conservativeness, limited R&D investment, and consequential insufficient technological capability, family firms concentrate on R&D projects to develop low-value technology that could be developed without major R&D investment. We call such a R&D resource allocation a low-value focused R&D strategy. Using such a strategy, family firms can produce more innovative output with less innovative input (R. C. Anderson, Duru, & Reeb, 2010; Block, Miller, Jaskiewicz, & Spiegel, 2013; De Massis, Frattini, Pizzurno, & Cassia, 2015).

On the other hand, there is another hypothetical explanation proposing that family firms show high innovation productivity not because they have no choice to focus on low-value innovation but because they have distinctive innovation strategies. An innovation strategy is constructed using three key drivers: innovation search, innovation process management, and innovation types (De Massis, Di Minin, & Frattini, 2015). Several family business researchers have found that incremental innovation through tradition and narrow and internal searches encapsulate family firms’ innovation strategies (Chirico & Salvato, 2008; Classen, Van Gils, Bammens, & Carree, 2012; De Massis, Frattini, Kotlar, Messeni Petruzelli, & Wright, 2016; De Massis et al., 2015). We consider such innovation strategies to be family firms’ signature moves, through which they avoid radical innovation and focus on incremental innovation.

According to scholarly discussion about innovation research, incremental innovation and low-value innovation seem similar but are rather different. Radical innovations are built on technology that a firm has not accumulated; as such, they are likely to change incumbent technological concepts and destroy existing competence (P. Anderson & Tushman, 1990; Das, 1994; Henderson & Clark, 1990; Rosenkopf & Nerkar, 2001). Innovation scholars consider that radical innovation must be novel in a sense that radicalness is a combination of elements and settings not previously observed (Dahlin & Behrens, 2005; Fleming, 2001; Hargadon & Sutton, 1997). The characteristics of radical innovations are contradictory to family firms’ signature moves stated above. On the other hand, incremental innovation has a cumulative nature and is developed based on existing technology, and its characteristics fit family firms’ signature moves. Therefore, family firms that employ such signature moves avoid developing radical innovations, focusing instead on incremental innovations. Due to their signature moves, family firms may convert lesser R&D into many incremental innovations, most of which are low-value innovations but some of which may be middle- or high-value innovations.

This study compares the R&D of family and nonfamily firms in the Japanese pharmaceutical industry. The pharmaceutical industry is a typical R&D-intensive industry in which innovation is crucial to firm performance; consequently, large pharmaceutical firms invest significant sums of money to develop blockbusters, products with exceptional sales much larger than the average. Since strong economies of scale in R&D exist in the industry, insufficient R&D investment by firms is considered fatal. Therefore, the pharmaceutical industry can be considered the most difficult environment for family firms with lower R&D intensity to compete. Nevertheless, many family firms survive in the Japanese pharmaceutical industry. Thus, the Japanese pharmaceutical industry fits the analytical characteristics we intend to examine.

Our analysis found that family firms show higher physical innovation productivity (the quantity of innovative output per R&D) than their nonfamily counterparts. However, there is no significant difference in average value of innovation (the value of innovative output per R&D) between family and nonfamily firms. Thus, these results do not suggest that family firms produce more innovations because they develop only low-value innovations. Instead, we found that family firms have a frequency distribution over the value of innovative output different from that of nonfamily firms. The distribution of nonfounder-managed family firms has a less positive skewness and a smaller kurtosis, indicating that the distribution has shorter and thinner right tail than the distribution of nonfamily firms. This suggests that nonfounder-managed family firms avoid developing radical innovations, focusing instead on incremental (nonradical) innovation by using their signature moves, innovation through tradition and narrow and internal search. Using baseball as a metaphor, it is as if they do not, as the idiom goes, swing for the fences to try to hit home runs (radical innovations); instead, they get many hits and, hence, reach bases by getting not only singles but also doubles and triples (nonradical innovations). We refer to this as a contact-hitting R&D strategy. Consequently, (nonfounder-managed) family firms with low R&D intensity can survive, even in an R&D-intensive industry, due to such a distinctive R&D strategy.

This study contributes to the literature on innovation and R&D strategy by family firms. While many previous studies have discussed why family firms with entrepreneurial orientation and long-time horizon invest less in R&D than their nonfamily counterparts (Chrisman, Chua, De Massis, Frattini, & Wright, 2015; De Massis et al., 2015) based on the ability and willingness framework, few studies have discussed how family firms can solve the ability–willingness paradox and build competitive advantage through innovation (Bigliardi, Cassia, De Massis, and Frattini, 2013; De Massis et al., 2015). This study, on the other hand, examines allocation of R&D resources among different projects (i.e., innovative outputs), while many previous studies have examined the amount of R&D investment (i.e., innovative inputs). We developed a model of contact-hitting R&D strategy and tested it by examining the frequency distribution over innovation value, where some previous studies have analyzed average value of innovation. By doing this, we attempt to contribute to a great understanding of family firms’ distinctive R&D strategy (R&D resource allocation) and innovation.

Literature Review and Hypothesis Development

Review of the Existing Studies on Family Firms’ Innovation

Many studies on innovation in family firms have been conducted over the past 20 years. Initially, most of these studies investigated if family firm characteristics support or impede entrepreneurial activities. While some studies argued that family firms are conservative and unwilling to engage in entrepreneurial activities (Sharma, Chrisman, & Chua, 1997), many other studies found that family firms have certain entrepreneurial characteristics such as a risk-taking nature and long-time horizons (James, 1999; Miller & Le Breton-Miller, 2005; Stein, 1988, 1989; Zahra, 2005; Zellweger, 2007).

These findings—that family firms display more entrepreneurial characteristics than nonfamily firms—are often explained by agency theory emphasizing conflicts between owners and managers of firms that function in a framework of separated ownership and management (Jensen, 1989; Jensen & Meckling, 1976; Morck, Shleifer, & Vishny, 1988). In an R&D context, managers tend to refrain from investing in risky R&D projects because of information asymmetry and moral hazard. However, the family owners, holding a large equity stake in the firm, are sufficiently incentivized to monitor and control managers. Moreover, agency conflicts are reduced due to alignment of interests between owners and managers in family firms owned and managed by the founding families (Claessens, Djankov, Fan, & Lang, 2002; Morck et al., 1988). Thus, family firms are considered to be more likely to engage in risky R&D projects than nonfamily firms (Block, 2012). Several researchers have found that family firms attach substantial importance to innovations and implement more innovations than their nonfamily counterparts (Classen, Carree, Van Gils, & Peters, 2014; Craig & Moores, 2006; Gudmundson, Tower, & Hartman, 2003; Simon, 1996).

Founding families may not only possess the incentives but also the capabilities to monitor managers effectively. The founders, their successors, and other family members share a lot of knowledge regarding their business, technologies, markets, the firm’s strategy, mission, shared values, and organizational culture (Le Breton-Miller & Miller, 2015). Hence, family managers with this abundant knowledge can better evaluate, accumulate, bundle, and leverage their resources. In other words, family firms are better at resource orchestration, which supports innovation (Sirmon & Hitt, 2003). Family firms also have a deeper internal search (or narrower search in terms of number of different external sources) than nonfamily firms (Chirico & Salvato, 2008; Classen et al., 2012; De Massis et al., 2016). Kotlar, De Massis, Frattini, Bianchi, and Fang (2013) found that family firms are generally reluctant to acquire external technology, unless the managers think that they can control over technological trajectory because of some protection mechanisms. Moreover, family firms try to keep a cohesive community of employees and have long-term, trust-based relationships with employees (De Massis, Audretsch, Uhlaner, & Kammerlander, 2018; Kammerlander & Van Essen, 2017; Miller & Le Breton-Miller, 2005). Therefore, family firms are more likely to adopt an autonomous decision process and less-formalized approach to governing the innovation process, which thereby promotes innovation (De Massis et al., 2015).

On the other hand, many researchers have examined whether any difference in R&D intensity exists between family and nonfamily firms. Most empirical studies have found that family firms invest less in R&D relative to nonfamily firms (Block, 2012; Block et al., 2011; Chen & Hsu, 2009; Chrisman & Patel, 2012; Munari et al., 2010), and therefore, it was accepted that family firms are conservative and risk-averse (Bloom & Van Reenen, 2007). There are several reasons for family firms’ lower R&D intensity. First, family owners do not want to dilute their equity or incur large amounts of debt due to their preference for maintaining a healthy balance sheet (Dreux, 1990; Romano, Tanewski, & Smyrnios, 2001; Ward, 2004). Therefore, they cannot raise enough funds to invest in R&D. Moreover, family owners’ wealth concentration in their firms makes them conservative and encourages them to avoid investing in risky R&D projects. These discussions are consistent with the theory of socioemotional wealth. The owners of family firms attempt to preserve socioemotional wealth, which includes the nonfinancial aspects of the firm that meet the family’s affective needs such as identity, the ability to exercise family influence, and the perpetuation of the family dynasty (Gómez-Mejía, Haynes, Núñez-Nickel, Jacobson, & Moyano-Fuentes, 2007). Since risky R&D projects may damage their socioemotional wealth, family firms tend to be conservative and invest less in R&D.

Several studies have explained the lower R&D intensity of family firms using the behavioral agency model (Chrisman & Patel, 2012; Gómez-Mejía et al., 2014; Patel & Chrisman, 2014). Firms make R&D investments by balancing potential gains and losses associated with them. Managers form reference points that reflect not only their economic goals but also their noneconomic goals; moreover, managers decide how much their firms will invest in R&D based on their reference points (Kotlar, Fang, De Massis, & Frattini, 2014). In the case of family firms, the managers take socioemotional wealth (family goals) into consideration in addition to the financial gain and loss (economic goals) (Kotlar, De Massis, Fang, & Frattini, 2014). A larger potential socioemotional loss associated with R&D investment relative to gain normally causes family firms invest less in R&D (Gómez-Mejía et al., 2014).

De Massis, Frattini, and Lichtenthaler (2013) reviewed prior studies finding mixed results on the relationship between family involvement and innovation using the framework of the three major steps of technological innovation: innovation input, activities, and output. As seen earlier, the findings on innovation input are largely consistent, suggesting a negative relationship between family involvement and investment in R&D. On the other hand, previous studies on innovation output found ambiguous and mixed results, as summarized above. Fewer studies exist on innovation activities than on input or output; however, such studies include the studies mentioned above on distinctive activities or processes such as control and monitoring, resource orchestration, search breadth, and the decision process.

To explain these seemingly contradictory findings that family involvement has both negative impacts on innovation inputs and positive impacts on innovation outputs, Chrisman et al. (2015) proposed a framework based on ability (discretion to act) and willingness (disposition to act).¹ They argued that the greater discretion that family owners are believed to have in combination with family firms’ unwillingness to innovate thus constitutes a paradox. Using this framework, Matzler, Veider, Hautz, and Stadler (2015) argued that family management and governance is negatively related to innovation input yet positively related to innovation output. Sciascia, Nordqvist, Mazzola, and De Massis (2015) suggested that the long-term orientation of family firms increases their ability to invest in R&D but not their willingness to do so.

Similarly, De Massis et al. (2015) argued that family firms have a dual nature as both conservative and innovative and that both of these aspects stand in tension owing to the fact that family firms face a paradox of less willingness to innovate while having more ability to do so. They suggested three contingency factors—(1) where: the direction in which the family firm wants to go (willingness), (2) how: the discretion the family firm has to move in that direction (ability as discretion), and (3) what: the resources and capabilities that are needed for the family firm to move in that direction (ability as resources).

Hypotheses on Innovation Productivity and Strategy of Family Firms

Previous studies on innovation in family firms have found mixed results on whether family firms are conservative or innovative. Moreover, they have found that family involvement has a negative impact on innovative inputs while it has a positive impact on innovative outputs. In fact, certain researchers have concluded that family firms have a dual nature when it comes to innovation (conservativeness vs. innovativeness) due to the paradoxical relationship between their willingness and ability to innovate—where, even as they are more able to, they are less willing to innovate. In this regard, a question remains: Given this paradox, how can family firms compete in industry where innovation is important for firm performance? Or put another way: How can family firms resolve the paradox and build competitive advantage through innovation? These questions were raised by Chrisman et al. (2015) and De Massis et al. (2015), respectively.

While few studies have tried to address the question, Chrisman and Patel (2012) and Patel and Chrisman (2014) discuss R&D strategy for family firms to survive in R&D intensive industry.² Chrisman and Patel (2012) found both that family firms invest less in R&D and also that the variability of their investment is greater relative to that of nonfamily firms when performance is above aspiration levels, whereas family firms increase R&D investments and decrease the variability of their investment when performance falls below aspiration levels. They argue that this is because family goals and economic goals diverge in the former situation, whereas they converge in the latter situation. Using a risk-abatement model, Patel and Chrisman (2014) found that family firms make more exploitative R&D investments resulting in more reliable and less risky sales levels when performance exceeds aspirations, whereas they make more exploratory R&D investments that may lead to higher but less reliable sales levels when performance falls below aspiration. However, an accumulation of technological knowledge, which can be a long process, is needed to develop innovations. Thus, it would be impossible for family firms to develop innovations and immediately recover their financial performance, even if a poorly performing family firm increased its R&D investment. Such firms would always have to continue to increase R&D. Therefore, previous studies on the R&D strategy of family firms do not fully explain how family firms can survive and compete in R&D-intensive industries.

Low-Value Focused R&D Strategy

There are, however, other studies that offer answers to the question. Chen and Hsu (2009) have suggested that family firms use their R&D resources more efficiently. As Ahuja, Lampert, and Tandon (2008) argued, it is not the level of innovative input but the process of converting input into output that matters, and if family firms achieve higher innovation performance with less innovative input (i.e., higher innovation productivity) than nonfamily firms, they can compete in R&D-intensive industries despite less R&D investment. Actually, several studies have found that family firms demonstrate higher physical innovation productivity (the amount of innovation per innovative input). For example, R. C. Anderson et al. (2010) have found that family firms produce more patent applications per R&D investment than their nonfamily counterparts. Then, the question is how family firms show higher physical innovation productivity than nonfamily firms.

Managers may allocate R&D resources between high-return and low-return R&D projects. High-return projects need huge R&D investment and high technological capabilities. As noted earlier, family firms invest less in R&D and consequently may have insufficient technological capabilities. Therefore, they may have to engage solely in low-return R&D projects. Moreover, high-return projects have a low likelihood of success, and low-return projects have a relatively high likelihood of success. According to the cumulative prospect theory, individuals tend to place too much weight on unlikely extreme outcomes (Barberis, 2013; Tversky & Kahneman, 1992). Similarly, in the context of R&D, external investors overemphasize projects producing high-value innovative output with low probabilities of success. Since nonfamily firms are owned by nonfamily shareholders and managed by professional managers, they prefer projects producing high-value innovative output with low probabilities of success. Family firms, meanwhile, are risk-averse because of wealth concentration and the fear of socioemotional wealth loss, as mentioned earlier (see Review of the Existing Studies on Family Firms’ Innovation).

Moreover, major R&D investment and advanced technological capability are needed to develop high-value innovation, while low-value innovations may be possible to develop without huge R&D investments or advanced technological capabilities. Therefore, family firms, which invest little in R&D and consequently do not have advanced technology, cannot develop high-value innovation, and instead have to deal with many R&D projects for low-value innovations. Moreover, low-value innovations have higher probabilities of success than high-value innovations, and hence, firms focusing their R&D on low-value innovations can produce more innovative outputs per innovative input (higher physical innovation productivity) while they can produce less valuable innovative outputs per innovative input (lower value innovation productivity) compared with their nonfamily counterparts (Lanjouw & Schankerman, 2004). In fact, R. C. Anderson et al. (2010) have found that family firms receive fewer citations (value of innovative output) per R&D (innovative input) than nonfamily firms, though family firms produce as many patents (number of innovation output) per R&D as nonfamily firms. Block et al. (2013) also found that given the firms’ level of R&D spending, family firms receive fewer citations. Their findings suggest that family firms may show higher physical innovative productivity because they are too conservative to develop high-value innovations; instead, they only develop low-value innovations with higher probabilities of success that require little R&D. We refer to this as the low-value R&D strategy hypothesis. If this hypothesis holds, we would see family firms develop many low-value innovations. Thus, we set the following hypothesis:

Hypothesis 1: Family firms show lower value innovation productivity (fewer citations per R&D) than nonfamily firms.

Contact-Hitting R&D Strategy

Managers may not allocate R&D resources between high return and low return projects. While we have thus far classified innovations based on their value, innovations can be also classified by their radicalness. Though value and radicalness are different factors, as Dahlin and Behrens (2005) noted, past innovation studies sometimes confound the radicalness of innovations with their commercial success (namely value) of innovations. In reality, not all high-value innovations are radical innovations; nor are all radical innovations high-value innovations. Existing innovation literature has studied the characteristics of radical innovations. Dewar and Dutton (1986) argue that firms need to process new information to develop radical innovations. P. Anderson and Tushman (1990) and Henderson (1993) characterize some radical innovation as competence-destroying. Henderson and Clark (1990) and Das (1994) argue that radical innovations change design architecture, core technology, and knowledge. Christensen and Rosenbloom (1995) and Christensen and Bower (1996) argue that radical innovations change the technological direction and performance trajectory. Rosenkopf and Nerkar (2001) argue that radical innovations are based on the technology the firm has not accumulated. Dahlin and Behrens (2005) state that radical innovation must be novel in the sense that radicalness is a combination of elements and settings not previously observed (Fleming, 2001; Hargadon & Sutton, 1997).

To develop radical innovations with the characteristics stated above, firms have to adopt a certain innovation strategy constructed by the three key drivers: innovation search (broad vs. narrow, exploration vs. exploitation), innovation process management (open vs. closed, external vs. internal), and innovation types (radical vs. incremental) (De Massis et al., 2015). Since radical innovation is novel, based on external and new technology, and tends to destroy existing systems and competencies, firms have to adopt broad and exploratory searches and open innovation by relying on external sources of knowledge.

On the other hand, family firms are good at an innovation strategy different from that which fits radical innovations. Classen et al. (2012) show that family small- and medium-sized enterprises (SMEs) have a lower search breadth than nonfamily SMEs, because family firms want to preserve their socioemotional wealth. As De Massis et al. (2016) suggested, family firms succeed in innovation by leveraging tradition to develop successful new products due to their strong links with the past. As stated above, family firms have a deeper internal search (or narrower search in terms of number of different external sources) than nonfamily firms (Chirico & Salvato, 2008; Classen et al., 2012; De Massis et al., 2016).³ Therefore, as Brinkerink (2018) summarizes, family firms are good at engaging effective deep internal searches and recombining new and existing knowledge assets. These are family firms’ distinctive innovation strategies and signature moves.

Since such signature moves are incompatible with radical innovations but are compatible with incremental innovations, family firms tend to avoid the former and focus on the latter.⁴ Actually, De Massis et al. (2015) found in a case study of 10 firms that family firms tend to focus their R&D on incremental innovations. If family firms avoid developing radical innovations and use their signature moves, they may be able to produce a greater amount of innovative outputs and show higher physical innovation productivity. In the pharmaceutical industry, blockbusters are usually very radical innovations. Thus, family pharmaceutical manufacturers may tend to avoid developing blockbusters (radical innovations) and focus their R&D on incremental innovations. However, the value of innovation is realized ex post. Most of the nonradical incremental innovations through tradition by deep internal search have only low and medium values, while some of them may be high value. A certain high-value innovation, which is radical for firms that have not accumulated the necessary technology to develop it, could be incremental for a firm that is familiar with the area of the innovation and has accumulated the necessary technology.

Figuratively speaking, radical innovations (blockbusters) are like homeruns in baseball—while they are valuable, their probability of success is quite low. Thus, firms developing blockbusters are power hitters who overswing (huge R&D investments) and hit homeruns (blockbusters) at the expense of batting average (efficiency). On the other hand, family firms that avoid developing radical innovations and focus on incremental innovations with their signature moves are contact hitters, who show high batting averages and get many hits, which result in singles and the occasional double or triple (nonradical innovations). We refer to such behavior of family firms as a contact-hitting R&D strategy.

It is assumed that in R&D-intensive industries, especially pharmaceuticals, enormous R&D investment is necessary to develop radical innovation. The competitive advantages of firms with high R&D intensity in the pharmaceutical industry can be generally explained as economies of scale in innovations with blockbusters. The cumulative distribution of returns from R&D projects is highly skewed to the right and has a fat right tail with an approximate power law structure $S^{- α}$ (Scherer, 2000; Scherer & Harhoff, 2000; Silverberg & Verspagen, 2007). That is, while most innovations yield modest returns, a few create high returns. In the pharmaceutical industry, for example, only a few drugs become blockbusters, and those help to cover the costs of the many low-value projects. Suppose companies A and B with size N_A and N_B merged into company M. If $α < 1$ , the likelihood of earning more than any fixed factor proportional to its size is larger for the merged company M of size N_A + N_B than for either of the two component companies. This is because a very large payoff from a successful project can pay for all the losing projects in situations where economies of scale occur.⁵

If firms try to develop blockbusters, the frequency distribution over the value of innovative output should have a long and thick right tail (Giuri et al., 2007; Morris, Teisberg, & Kolbe, 1991; Scherer & Harhoff, 2000; Silverberg & Verspagen, 2007; Wada, 2009). Since family and nonfamily firms have different preferences in terms of project risk and development targets, they should have different frequency distribution over the value of innovative output. Nonfamily firms—that prefer projects producing blockbusters, that is, radical and high-value innovations with a low likelihood of success—should have a distribution with a long and thick right tail, therefore showing that the distribution is highly positively skewed and has a larger kurtosis. Meanwhile, family firms—that avoid developing radical innovations and instead use their signature moves and focus on projects producing incremental innovations with a high likelihood of success—should have a less positively skewed distribution with a smaller kurtosis than nonfamily firms. Thus, we propose the following hypotheses corresponding to the contact-hitting R&D strategy hypothesis:

Hypothesis 2: Family firms have a frequency distribution over the value of innovative output with a shorter and thinner right tail than nonfamily firms have.

Data and Methods

Sample and Data Sources

In this study, we examined the R&D behavior of publicly traded firms in the Japanese pharmaceutical industry. The Japanese pharmaceutical market was the second largest in the world in 2013, and domestic firms dominated the relatively large market share.⁶ While a few “Mega Pharmas” exist, there are many relatively small manufacturers, around half of which are family firms. Japan has many unlisted small and medium-sized firms, most of which are family firms. However, we focused on publicly traded firms, partly because of the unavailability of data for unlisted companies. One advantage of this approach is that we can test our hypotheses by quasi-matching family firms with nonfamily firms on the basis that both are publicly traded. In this study, we used a set of panel data of publicly traded pharmaceutical firms in Japan from 1995 to 2007.

To prepare the sample, we first identified 46 firms that covered virtually all pharmaceutical firms active in the industry during the observed period based on the Nikkei Industry Code. We were able to identify the founding families of the 46 firms, regardless of their recent ownership structure.⁷ Then, we collected data on the ownership structure, top management, financial situation, and patents of each firm. To specify firm type (family vs. nonfamily), we collected data on the equity shares of the 10 largest shareholders from the Yuka Shoken Hokokusho (Japanese Form 10K) of each firm. We considered that the sum of the equity shares of founding family members was the family share.⁸ We also examined whether the president or chairperson was a family member by looking at the Yuka Shoken Hokokusho of each firm. Since the 10 largest shareholders and top management of a firm can change annually, the firm type (family or nonfamily firm) can change every year.

Among the 46 examined firms, we excluded one merged firm and three firms for which we could only find 1 year of data. Moreover, we excluded three firms that were outliers with more than 100% R&D intensity. Consequently, the sample comprised 39 firms. Financial data, including R&D expenditure, were obtained from the Nikkei Financial Quest database. As a result, we obtained unbalanced panel data consisting of 503 observations for the estimation of R&D intensity. Several firms went public during the observation period, and data were not available for the period before they were listed. Moreover, one firm became a holding company during the observation period, and data on subsidiaries since that time were not available. Theoretically, entry into (IPO) or exit from (delisting) the data set could be confounding factors for our analyses, since accessibility to the external financial market may be different between family and nonfamily firms (Crespí & Martín-Oliver, 2015). We believe, however, that such a potential sample selection problem does not nullify the results of this analysis since we have observations only when family firms were public (i.e., when the financial attributes of family firms become much closer to those of nonfamily firms). Patent data were obtained mainly from the Institute of Intellectual Property patent database and augmented with the EPO PATSTAT database for patent classifications. We collected data on the patents applied for by each of the sample firms by year from 1995 to 2009. Since the Institute of Intellectual Property patent database assigns only a single four-digit class to each patent based on the international patent classification, we obtained patent classifications from the PATSTAT database. We constructed the variables on patent applications using the two databases.

Variables and Methods

Our main independent variables are firm types. De Massis et al. (2013) emphasized that heterogeneity in family firms should be taken into account to better understand their behavior with regard to technological innovation. Thus, we have two family-managed firm variables since R&D decisions are made directly by the top management. The first family firm variable is a dummy variable (FBD_founder) for firms managed by the founder. FBD_founder is equal to 1 if the top management is the founder and 0 otherwise. The second family firm variable is a dummy variable (FBD_nonfounder) for firms managed by nonfounder family members. FBD_nonfounder is equal to 1 if the top management is not the founder but from the founding family and 0 otherwise. Since family owners have some influence on the decisions of top management, we controlled the effect of family ownership, including family share (Family Share) as a control variable.

Before testing the hypotheses, we examined the differences in R&D intensity and physical innovation productivity between family and nonfamily firms in the Japanese pharmaceutical industry to confirm the findings of previous studies. We have two measures for R&D intensity: logged R&D expenditure divided by sales (rdr) and logged R&D stock divided by sales (RDSTOCKr). Although previous studies have often used R&D expenditure sales ratio, single-year R&D expenditures may not be appropriate since patents, the measure of innovative output in this study, are the outcomes of cumulative R&D efforts. Therefore, we also constructed R&D stock (RDSTOCK) as follows, using the past history of R&D expenditure (RD) and the depreciation rate (Hall, Thoma, & Torrisi, 2007):

R D S T O C K_{t} = R D_{t} + (1 - δ) R D S T O C K_{t - 1},

where $δ$ is the depreciation rate. We followed conventions and used a 15% depreciation rate. We calculated the starting R&D stock at the first observation year as $R D S T O C K_{0} = R D_{0} / (δ + g),$ assuming that real R&D had been growing at a constant annual growth rate (g) prior to the sample. We used the 8% growth rate as applied in previous studies. If family firms show lower R&D intensity than nonfamily firms, the family firm variable should have a negative coefficient.

As to physical innovation productivity, we measured the number of patent applications per R&D expenditure (ap_num/RD)—that is, patent applications as innovative output and R&D expenditures as innovative input. If family firms use their R&D resources more efficiently than nonfamily firms, the former should produce more innovative output per R&D expenditure than the latter. Although previous studies have often used this variable, we also measured the number of patent applications per R&D stock (ap_num/RDSTOCK) for the reason stated above. If family firms show higher physical innovation productivity than nonfamily firms, the family firm variable should have a positive coefficient.

For Hypothesis 1 (the low-value focused R&D strategy), we examined whether family firms focus their R&D more on lower value innovation than nonfamily firms, and hence show lower value innovation productivity than nonfamily firms. We measured the number of forward citations (i.e., citations made by posterior patents) a firm received for its patents by application year as the value of innovative output; this was because, in previous studies, the economic value of innovation has often been measured by the number of forward citations a patent received (Block et al., 2013; Hall, Jaffe, & Trajtenberg, 2005; Trajtenberg, 1990). Consequently, we constructed two variables for value innovation productivity measures: the number of forward citations per R&D expenditure (citation/RD) and those per R&D stock (citation/RDSTOCK). If family firms adopt a low-value focused R&D strategy and focus their R&D on low-value innovation, the family firm variable should have a negative coefficient.

For the analyses concerning R&D intensity, physical innovation productivity, and value innovation productivity, we included many control variables in the panel analysis, following previous studies (e.g., Block, 2012; Block et al., 2011, 2013; Chen & Hsu, 2009). First, since economies of scale exist in the innovation process, we had to control for firm size. As a proxy for firm size, we employed the number of employees (emp). The expected sign of this variable is positive. We also controlled for firm age, which is the number of years from the foundation year to the observation year (age). Although older firms accumulate technological knowledge, which promotes the innovation process, they are often considered less innovative. Thus, we do not have any specific expectation about the sign of this variable. We included the debt–equity ratio (debtr) to control for the effect of debt levels. The expected sign of this variable is negative. As a variable of investment opportunities, we calculated the simple Tobin’s Q, defined as the sum of market capitalization and the book value of debt divided by the book value of total assets (q). The expected sign is positive. To control for the effects of cash in hand, we also included cash flow (the sum of net income and depreciation) divided by total assets (cf) (Brown & Petersen, 2011). All the above control variables were logged to reduce skewness.

The correlations and descriptive statistics of each variable in the sample are shown in Table 1. We estimated each of the dependent variables described earlier—rdr, RDSTOCKr, ap_num/RD, ap_num/RDSTOCK, citation/RD, citation/RDSTOCK—using fixed-effect linear regression, which Hausman tests indicated should be adopted.

Table 1.

Correlation Matrix and Descriptive Statistics (1).

		1	2	3	4	5	6	7	8	9	10	11	12	13	14
1	rdr	1
2	RDSTOCKr	0.85	1
3	ap_num/RD	−0.59	−0.58	1
4	ap_num/RDSTOCK	−0.49	−0.53	0.93	1
5	citation/RD	−0.47	−0.49	0.78	0.78	1
6	citation/RDSTOCK	−0.43	−0.48	0.78	0.85	0.96	1
7	emp	0.34	0.38	−0.32	−0.30	−0.24	−0.24	1
8	age	0.14	0.30	−0.27	−0.25	−0.23	−0.22	0.55	1
9	debtr	−0.44	−0.26	0.19	0.16	0.19	0.19	−0.14	−0.06	1
10	q	0.09	0.08	−0.11	−0.09	−0.07	−0.06	0.33	0.04	−0.13	1
11	cf	0.24	0.19	−0.12	−0.12	−0.11	−0.11	0.27	0.05	−0.13	0.18	1
12	Family Share	−0.29	−0.38	0.15	0.15	0.09	0.10	−0.35	−0.28	−0.06	−0.09	−0.12	1
13	FBD_founder	0.11	−0.05	0.01	0.03	−0.01	0.01	−0.46	−0.47	−0.15	0.03	−0.04	0.10	1
14	FBD_nonfounder	−0.17	−0.17	0.11	0.10	0.08	0.08	−0.01	0.02	0.08	0.06	0.03	0.52	−0.27	1
	Obs	553	553	485	485	529	529	564	573	564	534	543	517	521	521
	Mean	2.12	−1.34	0.01	1.68E-03	0.01	2.15E-03	7.01	4.00	4.10	0.35	2.27	10.26	0.07	0.51
	Standard Deviation	0.86	0.73	0.02	4.90E-03	0.03	0.01	1.10	0.44	0.87	0.40	1.77	13.56	0.25	0.50
	Minimum	−2.20	−4.44	0	0	0	0	4.68	0	2.25	−0.59	−2.53	0	0	0
	Maximum	3.41	−0.19	0.33	0.08	0.41	0.10	9.32	4.70	7.12	1.58	13.00	63.79	1	1

Next, for Hypothesis 2 (the contact-hitting R&S strategy), we did three tests to examine if family firms, which avoid developing radical innovations and focus on development of incremental innovation, have a frequency distribution over the value of innovative output with a shorter and thinner right tail than nonfamily firms. The first test is the Kolmogorov–Smirnov test using the patent level data. In this test, the null hypothesis is that the two cumulative distributions are the same (Durbin, 1973). We compared cumulative distributions over the value of innovative output between nonfamily firms and founder-managed family firms and those between nonfamily firms and nonfounder-managed family firms. If family firms’ distribution has a shorter and thinner right tail than nonfamily firms’ distribution, we would expect both rejection of the null hypothesis and that a significant difference where sample values in family firms are larger than those in nonfamily firms.

The second and the third tests are t tests for the mean difference. A frequency distribution over the value of innovative output with a long and thick right tail is highly positively skewed. If family firms’ distribution has a shorter and thinner right tail than nonfamily firms’ distribution, the former is less positively skewed than the latter. Similarly, the distribution with a long and thick tail have a large kurtosis. If family firms’ distribution has a shorter and thinner right tail than nonfamily firms’ distribution, the former has a smaller kurtosis than the latter. Then, we ran the two random effect regressions. For one regression, the dependent variable is skewness (Skew) and the independent variables are the two dummy variables, FBD_founder and FBD_nonfounder. For the other regression, the dependent variable is kurtosis (Kurtosis) and the independent variables are the same as before. We would expect the dummy variables have significantly negative coefficients. The correlations and descriptive statistics of each variable in the sample are shown in Table 2.

Table 2.

Correlation Matrix and Descriptive Statistics (2).

		1	2	3	4
1	Skew	1
2	Kurtosis	0.92	1
3	FBD_founder	−0.15	−0.11	1
4	FBD_nonfounder	−0.15	−0.13	−0.26	1
	Obs	433	433	521	521
	Mean	2.18	10.70	0.07	0.51
	Standard deviation	1.80	15.95	0.25	0.50
	Minimum	−0.71	1	0	0
	Maximum	11.28	129.19	1	1

Results

Before testing the hypotheses, we examined the differences in R&D intensity and physical innovation productivity between family and nonfamily firms in the pharmaceutical industry. The results of the analyses are shown in Table 3. The results of logged R&D expenditure sales ratio (rdr) are shown in Models (1) and (2) and those of logged R&D stock sales ratio (RDSTOCKr) are shown in Models (3) and (4). Models (1) and (3) include the control variables only, while the dummy variables of family-managed firm and the family share variable are added in Models (2) and (4). The dummy of founder-managed firm (FBD_founder) and that of nonfounder-managed family firm (FBD_nonfounder) are negative and significant at the 1% level in both Models (2) and (4). These results confirm the findings of previous studies that family firms have lower R&D intensity than their nonfamily counterparts. Significantly positive coefficients of family share (Family Share), on the other hand, suggest that R&D intensity becomes higher as family share increases. Among the control variables, firm size (emp) is significantly positive in Models (1) and (2), firm age (age) is significantly positive in all models, debt–equity ratio (debtr) is significantly positive, and cash flow divided by total assets (cf) is significantly negative only in Model (4). However, Tobin’s Q (q) are not significant in any models.

Table 3.

Fixed-Effects Regressions on R&D Intensity, Patent Applications, and Citations Per R&D.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)
	DV: rdr		DV: RDSTOCKr		DV: ap_num/RD		DV: ap_num/RDSTOCK		DV: citation/RD		DV: citation/RDSTOCK
emp	0.22* (0.13)	0.26* (0.14)	−0.07 (0.08)	0.02 (0.08)	−0.01 (0.01)	−0.01 (0.01)	−7.28E−04 (1.51E−03)	−1.30E−03 (1.52E−03)	0.01 (0.01)	0.01 (0.01)	1.95E−03 (1.89E−03)	1.42E−03(2.05E−03)
age	1.58*** (0.25)	1.80*** (0.31)	2.02*** (0.16)	2.31*** (0.18)	−0.05*** (0.01)	−0.05*** (0.02)	−0.01*** (3.14E−03)	−0.01** (3.47E−03)	−0.08*** (0.02)	−0.10*** (0.02)	−0.02*** (3.77E−03)	−0.02*** (4.69E−03)
debtr	−0.05 (0.05)	−0.01 (0.05)	0.03 (0.03)	0.07** (0.03)	−2.66E−03 (2.66E−03)	−1.22E−03 (2.66E−03)	−4.37E–04 (5.79E−04)	−2.06E–04 (5.86E−04)	−7.93E−04 (3.37E−03)	−2.12E−03 (3.90E−03)	6.54E-05 (6.92E−04)	−2.25E−04 (8.02E−04)
q	0.01 (0.05)	0.07 (0.06)	2.43E−03 (0.03)	0.05 (0.03)	−7.69E−04 (2.94E−03)	−5.69E−04 (2.99E−03)	−1.81E–04 (6.42E−04)	−1.69E–04 (6.57E−04)	−1.73E−03 (3.89E−03)	−1.49E−03 (4.30E−03)	−3.32E-04 (7.99E-04)	−3.58E−04 (8.86E−04)
cf	−0.04 (0.03)	−0.04 (0.03)	−0.02 (0.02)	−0.03* (0.01)	−4.67-E03** (1.96E−03)	−4.14-E03** (1.93E−03)	−1.31E−03*** (4.27E−04)	−1.22E−03*** (4.25E−04)	−2.00E-03 (1.90E−03)	−1.63E-03 (1.97E−03)	−6.02E-04 (3.91E−04)	−5.40E−04 (4.06E−04)
Family Share		0.02** (0.01)		0.03*** (0.01)		−6.05E−05 (4.57E-04)		−1.52E-05 (1.01E−04)		−9.78E−04 (7.29E-04)		−2.21E−04 (1.50E−04)
FBD_founder		−1.00*** (0.22)		−0.95*** (0.13)		0.02** (0.01)		4.22E−03* (2.38E−03)		0.01 (0.02)		3.35E−03 (3.63E−03)
FBD_nonfounder		−0.53*** (0.16)		−0.44*** (0.09)		0.03*** (0.01)		0.01*** (1.73E−03)		0.02 (0.01)		2.95E−03 (2.49E−03)
cons.	−5.49*** (1.50)	−6.66*** (1.80)	−9.03*** (0.95)	−10.98*** (1.07)	0.27*** (0.08)	0.25*** (0.09)	0.05*** (0.02)	0.05** (0.02)	0.24** (0.11)	0.36*** (0.13)	0.05** (0.02)	0.07*** (0.03)
N observations	503	480	503	480	451	451	451	451	481	458	481	458
R² within	0.11	0.17	0.31	0.42	0.05	0.09	0.06	0.08	0.08	0.09	0.07	0.08
R² between	0.00	0.01	0.00	0.01	0.08	0.11	0.07	0.11	1.70E−03	0.01	0.01	0.01
R² overall	0.04	0.01	0.05	0.02	0.08	0.09	0.07	0.08	0.02	0.02	0.03	0.03

Standard error is presented in parentheses. ^bAsterisks denote statistical significance at the 1% (***), 5% (**), and 10% (*) levels.

The results of the number of patent applications per R&D expenditure (ap_num/RD) are shown in Models (5) and (6), and those of the number of patent applications per R&D stock (ap_num/RDSTOCK) are shown in Models (7) and (8). In Model (6), both FBD_founder and FBD_nonfounder are positive and significant at 5% and 1% level, respectively. In Model (8), they are also positive and significant at 10% and 1% level, respectively. These results indicate that family-managed firms show higher physical innovation productivity, producing more patent applications per R&D expenditure or per R&D stock than nonfamily firms. Among the control variables, age and cf are significantly negative. However, the other control variables are not significant in any models.

The results of the analysis for Hypothesis 1 (the low-value focused R&D strategy) are shown in Models (9) through (12) in Table 3. The dependent variable is the number of forward citations per R&D expenditure (citation/RD) in Models (9) and (10) and those per R&D stock (citation/RDSTOCK) in Models (11) and (12). According to Table 3, either FBD_founder or FBD_nonfounder, is not significant. As to the control variables, only age is significantly negative in any models, while the other control variables are not significant. These results indicate that family firms receive as many citations per R&D expenditure (or R&D stock) as nonfamily firms; thus, Hypothesis 1 is not supported, and therefore, we consider that family firms do not adopt a low-value focused R&D strategy of focusing their R&D on low-value innovation.

Next, we compared the frequency distribution over the value of patents between family and nonfamily firms (Hypothesis 2). The results of the Kolmogorov–Smirnov test are shown in Table 4. Since the p values in “Combined K-S” are very small, the distribution over the value of patents between family and nonfamily firms are different. The small p values of the second row in each comparison indicate that the sample values in nonfamily firms are significantly smaller than those in founder-managed family firms or nonfounder-managed family firms. These results suggest that the frequency distribution over the value of patents of family firms have a shorter and thinner right tail than that of nonfamily firms.

Table 4.

Kolmogorov-Smirnov test.

Founder-managed family firm versus nonfamily firm	Distance^a	p value
Founder < Nonfamily^b	0.02	.86
Founder > Nonfamily	−0.12	.00
Combined K-S:	0.12	.00
Nonfounder-managed family firm vs. Nonfamily firm	Distance	p value
Nonfounder < Nonfamily	0.00	.94
Nonfounder > Nonfamily	−0.10	.00
Combined K-S:	0.10	.00

Note. K-S = Kolmogorov–Smirnov test.

Distance is the maximum distance between the two cumulative distribution. ^b“Founder < Nonfamily” means that sample values in founder-managed family firms are smaller than those in nonfamily firms, and vice versa.

Moreover, as indicated in Table 5, (nonfounder-managed) family firms have a significantly different frequency distribution over the value of patents from that of nonfamily firms in terms of its skewness and kurtosis. Model (1) shows that both dummy variables are negative but only FBD_nonfounder is significant. It suggests that the frequency distribution over the value of patents of nonfounder-managed family firms are less positively skewed than that of nonfamily firms. Model (2) also indicates that FBD_nonfounder is significantly negative, meaning that the kurtosis of the frequency distribution over the value of patents of nonfounder-managed family firms is significantly smaller than that of nonfamily firms. These results suggest both that the frequency distribution over the value of patents of nonfounder-managed family firms has a shorter and thinner right tail than that of nonfamily firms, and that nonfounder-managed family firms avoid developing blockbusters (radical and high-value innovations) but focus on developing incremental innovations.

Table 5.

Random-Effects Regressions on Skewness and Kurtosis.

	(1)	(2)
	DV: Skew	DV: Kurtosis
FBD_founder	−0.83102 (0.65)	−6.56 (5.63)
FBD_nonfounder	−0.73* (0.38)	−5.86* (3.11)
Constant	2.43*** (0.31)	13.31*** (2.45)
N observations	418	418
R² within	2.90E−03	4.00E−03
R² between	0.07	0.0548
R² overall	0.05	0.0345

Note. DV = dependent variable.

Standard error is presented in parentheses. ^bAsterisks denote statistical significance at the 1% (***), 5% (**), and 10% (*) levels.

Discussion and Conclusion

This study examined why family firms show higher physical innovation productivity, analyzing data on the Japanese pharmaceutical industry. It proceeded from the supposition that family firms with low R&D intensity can compete in an R&D-intensive industry because of their higher physical innovation productivity. We confirmed that even in the R&D-intensive pharmaceutical industry, family-managed firms invest less in R&D and show higher physical innovation productivity—that is, they produce more patent applications per R&D than nonfamily firms. While we did not find any significant difference in value innovation productivity (the number of citations per R&D) between family and nonfamily firms, we did find that the frequency distributions over the value of patents are different between family and nonfamily firms by the Kolmogorov–Smirnov test, and that the frequency distribution over the value of patents of nonfounder-managed family firms has a less positive skewness and a smaller kurtosis and, hence, a shorter and thinner right tail than nonfamily firms.

These results suggest that family-managed firms show higher physical innovation productivity not because they develop only low-value innovations due to their insufficient technological capabilities or conservativeness (the low-value focused R&D strategy hypothesis). Instead, their innovation strategies, such as narrow and internal search and recombining new and existing knowledge assets, increase physical innovation productivity. Such innovation strategies are family firms’ signature moves, which do not fit developing radical innovations or blockbusters but do enhance incremental innovation. A distribution with a shorter and thinner right tail of nonfounder-managed family firms indicates that they avoid developing blockbusters (radical innovation) and allocate more R&D resources to many projects to develop incremental innovation (a contact-hitting R&D strategy).

Though we did not find supporting evidence for the low-value focused R&D strategy hypothesis and did find supporting evidence for the contact hitting R&D strategy, these two R&D strategies seem similar in that many incremental innovations are relatively low-value innovations. However, the results of this study may produce widely varying implications. The scholars who believe that family firms adopt the low-value focused R&D strategy tend to argue that family firms are less innovative and more apt to underperform in research-intensive sectors (Block et al., 2013). They argue that family firms should recognize that they are biased toward low-value innovation rather than radical innovation and change their R&D behavior by benchmarking their unbiased rivals and emphasizing radical innovation development to improve their performance in R&D-intensive industries. However, correcting their bias and changing their R&D behavior destroy the fit among the characteristics of family firms, the factors of innovation strategy and the type of innovation, and therefore, may have a deteriorating effect on their innovation performance. De Massis et al. (2015) suggest that creating fit among the key drivers of heterogeneity of family firms and innovation decision is vital to resolving the paradox in family firm innovation.

This study found Japanese family pharmaceutical firms adopt the contact-hitting R&D strategy and show as much high-value innovation productivity as their nonfamily counterparts. Even in this typical R&D intensive industry, where firms are supposed to develop one radical innovation after another, family firms can survive the adoption of the contact-hitting R&D strategy. Family firms develop many incremental innovations using their signature moves, even though they do not develop radical innovations because doing so is incompatible with their signature moves. Japanese pharmaceutical firms achieve fit among the characteristics of family firms, the factors of innovation strategy, and the type of innovation.

One example is Ono Pharmaceutical, which is known for its cancer treatment drug, Opdivo, to which the research of Nobel Prize winner Dr. Tasuku Honjo contributed. Ono Pharmaceutical used to be a family-managed firm when the company started its R&D of prostaglandin, a group of physiologically active lipid compounds, in 1965. Other nonfamily large pharmaceutical firms had also been developing prostaglandin and several other development targets simultaneously, but Ono focused its R&D on prostaglandin. Moreover, Ono did not conduct research regarding the circulatory and digestive systems, which were expected to have huge consumer demand; instead, they researched a niche area of pharmaceutical development—the urinary system. In so doing, it was able to develop several prostaglandin formulations sequentially (Takahashi, 2009).

Another example is Shionogi, a successful Japanese pharmaceutical firm managed by the founding family. While many pharmaceutical firms acquire new drugs from start-ups and foreign firms through license agreements, Shionogi develops new drugs in-house. Moreover, the company focuses its R&D on drugs for infectious diseases, a practice that is expected to have smaller demand. This has resulted in the company successfully developing a first-in-class anti-influenza medicine (Nikkei Business Daily, 2018). Based on these examples, we argue that it is more important for family firms to find a unique market in which they can apply their signature moves and invest their resources in R&D in order to fill that niche.

This study contributes to the literature on family firm innovation. We tried to distinguish among the reasons for family firms’ higher physical innovation productivity compared to their nonfamily counterparts. We tested and rejected the low-value focused R&D hypothesis, which said that family firms are biased toward low-value innovation. This is theory testing, according to the taxonomy of theoretical contributions in Colquitt and Zapata-Phelan (2007). Moreover, combining the argument on radical and incremental innovations by innovation scholars (Dahlin & Behrens, 2005; Fleming, 2001; Hargadon & Sutton, 1997) and family business innovation studies (Brinkerink, 2018; Classen et al., 2012; De Massis et al., 2016), we proposed that family firms adopting a contact-hitting R&D strategy. Family firms adopt a contact-hitting R&D strategy to develop many incremental innovations, which result in not only many low-value innovations but also a few mid- or high-value innovations. Therefore, family firms investing less in R&D can compete even in an R&D intensive industry, and a contact-hitting R&D strategy is one of the ways for family firms to resolve the ability–willingness paradox and reconcile their innovativeness with their risk aversion (Chrisman et al., 2015; De Massis et al., 2015). In this way, our study makes a theoretical contribution to the literature by constructing a new theory as well (Colquitt & Zapata-Phelan, 2007).

This research also contributes to the empirical study of the innovation of family firms and shows the importance of analyzing their innovative output. Most empirical studies of family-firm innovation have analyzed their innovative input. However, by analyzing the innovative output (i.e., product innovation) of several firms, De Massis et al. (2015) were able to demonstrate that family firms are engaged in incremental product innovation. This encouraged scholars to examine family firms’ innovation strategies statistically; the present study is one such statistical analysis of innovative output through patent data. Though other studies have also analyzed patent data, they examined the number of patent applications or patent citations and their ratio to R&D expenditure without considering right-tail distribution. To the best of our knowledge, no studies in the area of strategy or family business research have examined the asymmetry or tail-heaviness of frequency distribution over the value of patents. In pharmaceutical or other research-intensive industries where R&D and innovation play crucial roles, most firms, including family ones, develop not only low-value innovations but also mid- or even high-value innovations. Since skewness and kurtosis capture the tail behavior as well as the asymmetry of a density function (Patil, Patil, & Bagkavos, 2012), it is important to analyze not only the average value of patents but also the frequency distribution over the value of patents.

There are, of course, several limitations to this study. We tried to distinguish among the two hypotheses regarding why family firms show higher physical innovation productivity: the low-value focused R&D strategy and the contact-hitting R&D strategy. However, the hypotheses were not tested directly by analyzing organizational characteristics, resource allocation, or innovation management of family firms; rather, they were tested indirectly by analyzing their innovative output (patents) from various angles. It is expected that additional studies will directly test the hypotheses on innovation strategies of family firms using the data on innovation process. Second, the sample used in this study was limited to Japanese pharmaceutical firms. Thus, the generalizability of our findings may be limited. For example, large pharmaceutical firms in Japan are relatively small compared to leading pharmaceutical firms worldwide. There are several large pharmaceutical firms owned or managed by their founding families in the United States and Europe. Therefore, if we compare family and nonfamily firms throughout the world, where size differences among the firms are significant, we might find results that differ from those of this study. It would be beneficial to conduct additional studies on the innovation and R&D strategies of family firms that consider the same research question in different industrial and national contexts. Nevertheless, this paper fills in some of the research gaps and contributes to the field of research on innovation in family businesses, even though it considers only a single industry in a single country.

Footnotes

Acknowledgements

The authors would like to thank the associate editor of FBR, Alfredo De Massis, and the two anonymous reviewers for their helpful and constructive comments. The authors also thank the participants at the Workshop on Family Firm Management Research 2014 and 2016, the Asia Academy of Management Conference 2015, and the Academy of Management Annual Meeting 2016.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was financially supported by the grants (No. 25245052, 17H02570, and 18H00889) of Japan Society for the Promotion of Science.

Notes

ORCID iDs

Shigeru Asaba

Tetsuo Wada

Author Biographies

Shigeru Asaba is a professor of strategy in Waseda Business School at Waseda University. He received his PhD, in management from the University of California, Los Angeles. His current research focuses on distinctive strategies and business succession of family firms, competition and collaboration among firms, and change and creation of business models.

Tetsuo Wada is a professor at the Faculty of Economics, Gakushuin University. He has research interests in the process of creating and trading intellectual property rights. He obtained PhD in business administration from the University of California, Berkeley and LLB from the University of Tokyo.

References

Ahuja

Lampert

C. M.

Tandon

(2008). Moving beyond Schumpeter: Management research on the determinants of technological innovation. Academy of Management Annals, 2, 1-98.

Anderson

R. C.

Duru

Reeb

(2010). Family preferences and investment policy: Evidence from R&D spending and capital expenditures. Retrieved from https://www.researchgate.net/publication/252258825_Family_Preferences_and_Investment_Policy_Evidence_from_Capital_Expenditures_and_RD_Spending

Anderson

R. C.

Reeb

D. M.

(2003). Founding-family ownership and firm performance: Evidence from the S&P 500. Journal of Finance, 58, 1301-1327.

Anderson

Tushman

M. L.

(1990). Technological discontinuities and dominant designs: A cyclical model of technological change. Administrative Science Quarterly, 35, 604-633.

Arzubiaga

Kotlar

De Massis

Maseda

Iturralde

(2018). Entrepreneurial orientation and innovation in family SMEs: Unveiling the (actual) impact of the Board of Directors. Journal of Business Venturing, 33, 455-469.

Barberis

N. C.

(2013). Thirty years of prospect theory in economics: A review and assessment. Journal of Economic Perspectives, 27, 173-196.

Bigliardi

Cassia

De Massis

Frattini

(2013). Technology strategy in family business: A new avenue for research. International Journal of Entrepreneurship and Innovation Management, 17, 1-7.

Block

J. H.

(2012). R&D investments in family and founder firms: An agency perspective. Journal of Business Venturing, 27, 248-265.

Block

J. H.

Miller

Jaskiewicz

Spiegel

(2011). Innovation in founder and family firms: Entrepreneurial versus nurturer identities of owners. Frontiers of Entrepreneurship Research, 31, 437-450.

10.

Block

J. H.

Miller

Jaskiewicz

Spiegel

(2013). Economic and technological importance of innovations in large family and founder firms: An analysis of patent data. Family Business Review, 26, 180-199.

11.

Bloom

Van Reenen

(2007). Measuring and explaining management practices across firms and countries. Quarterly Journal of Economics, 122, 1351-1408.

12.

Brinkerink

(2018). Broad search, deep search, and the absorptive capacity performance of family and nonfamily firm R&D. Family Business Review, 31, 295-317. doi:10.1177/0894486518775187

13.

Brown

J. R.

Petersen

B. C.

(2011). Case holdings and R&D smoothing. Journal of Corporate Finance, 17, 694-709.

14.

Bozec

Di Vito

(2019). Founder-controlled firms and R&D investments: New evidence from Canada. Family Business Review, 32, 76-96.

15.

Chen

H. L.

Hsu

W. T.

(2009). Family ownership, board independence, and R&D investment. Family Business Review, 22, 347-362.

16.

Chirico

Salvato

(2008). Knowledge integration and dynamic organizational adaptation in family firms. Family Business Review, 21, 169-181.

17.

Chrisman

J. J.

Chua

J. H.

De Massis

Frattini

Wright

(2015). The ability and willingness paradox in family firm innovation. Journal of Product Innovation Management, 32, 310-318.

18.

Chrisman

J. J.

Patel

P. C.

(2012). Variations in R&D investments of family and nonfamily firms: Behavioral agency and myopic loss aversion perspectives. Academy of Management Journal, 55, 976-997.

19.

Christensen

C. M.

Bower

J. L.

(1996). Customer power, strategic investment, and the failure of leading firms. Strategic Management Journal, 17, 197-218.

20.

Christensen

C. M.

Rosenbloom

R. S.

(1995). Explaining the attacker’s advantage: Technological paradigms, organizational dynamics, and the value network. Research Policy, 24, 233-257.

21.

Classen

Carree

Van Gils

Peters

(2014). Innovation in family and non-family SMEs: An exploratory analysis. Small Business Economics, 42, 595-609.

22.

Classen

Van Gils

Bammens

Carree

(2012). Accessing resources from innovation partners: The search breadth of family SMEs. Journal of Small Business Management, 50, 191-215.

23.

Claessens

Djankov

Fan

J. P.

Lang

L. H.

(2002). Disentangling the incentive and entrenchment effects of large shareholdings. Journal of Finance, 57, 2741-2771.

24.

Colquitt

J. A.

Zapata-Phelan

C. P.

(2007). Trends in theory building and theory testing: A five-decade study of the Academy of Management Journal. Academy of Management Journal, 50, 1281-1303

25.

Craig

J. B. L.

Moores

(2006). A 10-year longitudinal investigation of strategy, systems, and environment on innovation in family firms. Family Business Review, 19, 1-10.

26.

Crespí

Martín-Oliver

(2015). Do family firms have better access to external finance during crises? Corporate Governance: An International Review, 23, 249-265.

27.

Dahlin

K. B.

Behrens

D. M.

(2005). When is an invention really radical? Defining and measuring technological radicalness. Research Policy, 34, 717-737.

28.

Das

(1994). A question of wines and bottles? Examining new technology and competitiveness between new technologies. Academy of Management Annual Meeting Proceedings, 1994(1), 335-339. doi:10.5465/AMBPP.1994.10345879

29.

De Massis

Audretsch

Uhlaner

Kammerlander

. (2018). Innovation with limited resources: Management lessons from the German Mittelstand. Journal of Product Innovation Management, 35, 125-146.

30.

De Massis

Di Minin

Frattini

. (2015). Family-driven innovation: Resolving the paradox in family firms. California Management Review, 58, 5-19.

31.

De Massis

Frattini

Kotlar

Messeni Petruzzelli

Wright

. (2016). Innovation through tradition: Lessons from innovative family businesses and directions for future research. Academy of Management Perspectives, 30, 93-116.

32.

De Massis

Frattini

Lichtenthaler

. (2013). Research on technological innovation in family firms present debates and future directions. Family Business Review, 26, 10-31.

33.

De Massis

Frattini

Pizzurno

Cassia

. (2015). Product innovation in family versus nonfamily firms: An exploratory analysis. Journal of Small Business Management, 53, 1-36.

34.

Dewar

Dutton

(1986). The adoption of radical and incremental innovations: An empirical analysis. Management Science, 32, 1422-1433.

35.

Dreux

D. R.

(1990). Financing family business: Alternatives to selling out or going public. Family Business Review, 3, 225-243.

36.

Duran

Kammerlander

Van Essen

Zellweger

(2016). Doing more with less: Innovation input and output in family firms. Academy of Management Journal, 59, 1224-1264.

37.

Durbin

(1973). Distribution theory for tests based on the sample distribution function (Vol. 9). Philadelphia, PA: Siam.

38.

Ebihara

Kubota

Takehara

Yokota

(2013). Database construction for family business research. Journal of Musashi University, 61, 49-57.

39.

Fang

Kotlar

Memili

Chrisman

J. J.

De Massis

(2018). The pursuit of international opportunities in family firms: Generational differences and the role of knowledge-based resources. Global Strategy Journal, 8, 136-157.

40.

Feranita

Kotlar

De Massis

(2017). Collaborative innovation in family firms: Past research, current debates and agenda for future research. Journal of Family Business Strategy, 8, 137-156.

41.

Fleming

(2001). Recombinant uncertainty in technological search. Management Science, 47, 117-132.

42.

Giuri

Mariani

Brusoni

Crespi

Francoz

Gambardella

. . . Verspagen

(2007). Inventors and invention processes in Europe: Results from the PatVal-EU Survey. Research Policy, 36, 1107-1127.

43.

Gómez-Mejía

L. R.

Campbell

J. T.

Martin

Hoskisson

R. E.

Makri

Sirmon

D. G.

(2014). Socioemotional wealth as a mixed gamble: Revisiting family firm R&D investments with the behavioral agency model. Entrepreneurship Theory and Practice, 38, 1351-1374.

44.

Gómez-Mejía

L. R.

Haynes

K. T.

Núñez-Nickel

Jacobson

K. J.

Moyano-Fuentes

(2007). Socioemotional wealth and business risks in family-controlled firms: Evidence from Spanish olive oil mills. Administrative Science Quarterly, 52, 106-137.

45.

Gudmundson

Tower

C. B.

Hartman

E. A.

(2003). Innovation in small businesses: Culture and ownership structure do matter. Journal of Developmental entrepreneurship, 8, 1-17.

46.

Hall

B. H.

Jaffe

Trajtenberg

(2005). Market value and patent citations. RAND Journal of Economics, 36, 16-38.

47.

Hall

Thoma

Torrisi

(2007). The market value of patents and R&D: Evidence from European firms (NBER Working Paper No. 13426). Cambridge, MA: National Bureau of Economic Research.

48.

Hargadon

Sutton

R. I.

(1997). Technology brokering and innovation in a product development firm. Administrative Science Quarterly, 42, 716-749.

49.

Henderson

(1993). Underinvestment and incompetence as responses to radical innovation: Evidence from the photolithographic alignment equipment industry. RAND Journal of Economics, 24, 248-270.

50.

Henderson

R. M.

Clark

K. B.

(1990). Architectural innovation: The reconfiguration of existing product technologies and the failure of established firms. Administrative Science Quarterly, 35, 9-30.

51.

James

H. S.

(1999). Owner as manager, extended horizons and the family firm. International Journal of the Economics of Business, 6. 41-55.

52.

Jensen

M. C.

(1989). Eclipse of the public corporation. Harvard Business Review, 67, 61-74.

53.

Jensen

M. C.

Meckling

W. H.

(1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3, 305-360.

54.

Kammerlander

Ganter

(2015). An attention-based view of family firm adaptation to discontinuous technological change: Exploring the role of family CEOs’ noneconomic goals. Journal of Product Innovation Management, 32, 361-383.

55.

Kammerlander

Van Essen

(2017). Research: Family firms are more innovative than other companies. Retrieved from https://hbr.org/2017/01/research-family-firms-are-more-innovative-than-other-companies

56.

Kotlar

De Massis

Fang

Frattini

(2014). Strategic reference points in family firms. Small Business Economics, 43, 597-619.

57.

Kotlar

De Massis

Frattini

Bianchi

Fang

(2013). Technology acquisition in family and nonfamily firms: A longitudinal analysis of Spanish manufacturing firms. Journal of Product Innovation Management, 30, 1073-1088.

58.

Kotlar

Fang

De Massis

Frattini

(2014). Profitability goals, control goals, and the R & D investment decisions of family and nonfamily firms. Journal of Product Innovation Management, 31, 1128-1145.

59.

Lanjouw

J. O.

Schankerman

(2004). Patent quality and research productivity: measuring innovation with multiple indicators. Economic Journal, 114, 441-465.

60.

Le Breton-Miller

Miller

. (2015). Learning stewardship in family firms: For family, by family, across the life cycle. Academy of Management Learning & Education, 14, 386-399.

61.

Matzler

Veider

Hautz

Stadler

(2015). The impact of family ownership, management, and governance on innovation. Journal of Product Innovation Management, 32, 319-333.

62.

Mazzelli

Kotlar

De Massis

(2018). Blending in while standing out: Selective conformity and new product introduction in family firms. Entrepreneurship Theory and Practice, 42, 206-230.

63.

Miller

Le Breton-Miller

(2005). Managing for the long run. Boston, MA: Harvard Business School Press.

64.

Miller

Le Breton-Miller

Lester

R. H.

Cannella

A. A.

Jr. (2007). Are family firms really superior performers? Journal of Corporate Finance, 13, 829-858.

65.

Morck

Shleifer

Vishny

R. W.

(1988). Management ownership and market valuation: An empirical analysis. Journal of Financial Economics, 20, 293-315.

66.

Morris

P. A.

Teisberg

E. O.

Kolbe

A. L.

(1991). When choosing R&D projects, go with long shots. Research Technology Management, 34, 35-40.

67.

Munari

Oriani

Sobrero

(2010). The effects of owner identity and external governance systems on R&D investments: A study of Western European firms. Research Policy, 39, 1093-1104.

68.

Nikkei Business Daily. (2018, October 17). The power structure of anti-influenza medicines will change completely.

69.

Patel

P. C.

Chrisman

J. J.

(2014). Risk abatement as a strategy for R&D investments in family firms. Strategic Management Journal, 35, 617-627.

70.

Patil

P. N.

Patil

P. P.

Bagkavos

(2012). A measure of asymmetry. Statistical Papers, 53, 971-985.

71.

Porta

Lopez-De-Silanes

Shleifer

(1999). Corporate ownership around the world. Journal of Finance, 54, 471-517.

72.

Romano

C. A.

Tanewski

G. A.

Smyrnios

K. X.

(2001). Capital structure decision making: A model for family business. Journal of Business Venturing, 16, 285-310.

73.

Rosenkopf

Nerkar

(2001). Beyond local search: Boundary-spanning, exploration, and impact in the optical disk industry. Strategic Management Journal, 22, 287-306.

74.

Scherer

F. M.

(2000). The size distribution of profits from innovation. In Encaoua

Hall

B. H.

Laisney

Mairesse

(Eds.), The economics and econometrics of innovation (pp. 473-494). New York, NY: Springer.

75.

Scherer

F. M.

Harhoff

(2000). Technology policy for a world of skew-distributed outcomes. Research Policy, 29. 559-566.

76.

Sciascia

Nordqvist

Mazzola

De Massis

(2015). Family ownership and R&D intensity in small- and medium-sized firms. Journal of Product Innovation Management, 32, 349-360.

77.

Sharma

Chrisman

J. J.

Chua

J. H.

(1997). Strategic management of the family business: Past research and future challenges. Family Business Review, 10, 1-35.

78.

Schumpeter

J. A.

(1926). Theorie der wirtschaftlichen Entwicklung: Eine Untersuchung über Unternehmergewinn, Kapital, Kredit, Zins und den Konjunkturzyklus [Theory of economic development: A study on entrepreneurial profit, capital, credit, interest rates and the business cycle]. Munich, Germany: Dunker und Humblot.

79.

Silverberg

Verspagen

(2007). The size distribution of innovations revisited: An application of extreme value statistics to citation and value measures of patent significance. Journal of Econometrics, 139, 318-339.

80.

Simon

(1996). Hidden champions: Lessons from 500 of the world’s best unknown companies. Cambridge, MA: Harvard Business Press.

81.

Sirmon

D. G.

Hitt

M. A.

(2003). Managing resources: Linking unique resources, management, and wealth creation in family firms. Entrepreneurship Theory and Practice, 27, 339-358.

82.

Solow

R. M.

(1957). Technical change and the aggregate production function. Review of Economics and Statistics, 39, 312-320.

83.

Sornette

(2002). Economies of scale in innovations with block-busters. Quantitative Finance, 2, 224-227.

84.

Stein

J. C.

(1988). Takeover threats and managerial myopia. Journal of Political Economy, 96, 61-80.

85.

Stein

J. C.

(1989). Efficient capital markets, inefficient firms: A model of myopic corporate behavior. Quarterly Journal of Economics, 104, 655-669.

86.

Takahashi

(2009). The growth of firms through resource accumulation and sequential niche entry: Case studies of R&D conducted for prostaglandin in Ono Pharmaceutical company. Journal of Business Management, 24, 41-53. (in Japanese)

87.

Trajtenberg

(1990). A penny for your quotes: Patent citations and the value of innovations. Rand Journal of Economics, 21, 172-187.

88.

Tversky

Kahneman

(1992). Advances in prospect theory: Cumulative representation of uncertainty. Journal of Risk and Uncertainty, 5, 297-323.

89.

Villalonga

Amit

(2006). How do family ownership, control and management affect firm value? Journal of Financial Economics, 80, 385-417.

90.

Wada

(2009). Backward patent citations and inventors’ recognition of differential influences (RIETI Discussion Paper Series 10-J-001). Retrieved from https://www.researchgate.net/publication/254412808_Backward_Patent_Citations_and_Inventors’_Recognition_of_Differential_Influences_Japanese

91.

Ward

J. L.

(2004). Perpetuating the family business: 50 lessons learned from long-lasting, successful families in business. New York, NY: Palgrave Macmillan.

92.

Zahra

S. A.

(2005). Entrepreneurial risk taking in family firms. Family Business Review, 18, 23-40.

93.

Zellweger

(2007). Time horizon, costs of equity capital, and generic investment strategies of firms. Family Business Review, 20, 1-15.