Entrepreneurship and dynamic capabilities: how firm age and size affect the ‘capability enhancement–SME performance’ relationship

The terms ‘OC’ and ‘DC’ refer to two different capabilities. To be clear: all firms have OCs in order to operate; not all firms have DCs, and fewer have highly-routinized DCs. All DCs change OCs.

The firm characteristics of newness and smallness are also attractive as empirical measures. They are relatively objective measures, simple to understand, and available. Each has been shown to affect firm performance in the entrepreneurship literature (i.e., these items are often used as controls; Steffens et al. 2009). These factors also provide a solid basis for building upon in future work because they proxy for more sophisticated factors such as: scale economies, market power, bargaining power, resource slack, specialization, experience, and so on.

Note in Table 1 that several descriptive results were sensible (e.g., the main reason given for not having a routinized DC was logical), and that the sample provided a good proportion of firms reporting DCs.

The final responses were the proxy for non-respondents and the early responses were the proxy for respondents. The t tests for differences between the two groups yielded no statistically significance in the survey items used in the analysis. We determined that there were no statistically significant differences between early and late respondents at the two-tailed p < 0.05 levels for any of the dependent or independent variables.

The procedural methods we used included:

The statistical methods we used included:

The Chronbach alpha is a coefficient of reliability or internal consistency of a construct, increasing with the inter-correlations among construct items. It is widely accepted as an indicator of the degree to which a set of items measures a single latent construct, where those items measure different substantive areas within that construct. The normal cutoff level for acceptability of reliability is an alpha of 0.70. The constructs are continuous variables created by combining—usually by taking the mean of—several separate Likert-scale survey question responses related to a specific measure.

The full survey upon which this current paper is based contained several other questions, many of which are complementary, but not focal to this study. For example, other questions related to the DC-building process, DC characteristics, DC use, and so on. We do not report these results or variables formally here because they are not focal to the points made in this paper.

The probit analysis was run using the following instruments: firm age and firm size—to control for experience and need related to building a DC (Pavlou and El Sawy 2011; Protogerou et al. 2012); industry and geography/location dummy variables—to control for contextual pressures and effects on DC attainment (Capron and Mitchell 2009; Lichtenthaler 2009); entrepreneurial orientation—to control for decision-making syle (e.g., risk-taking) on choosing a DC to enhance adaptability (Lichtenthaler 2009; Rothaermal and Hess 2007); firm resources—to control for how comfortably a firm could afford to build and maintain a DC (McKelvie and Davidsson 2009; Pavlou and El Sawy 2011); industry hostility and industry turbulence—to control for how the competitive pressures, the dynamism and unpredictability of the context could motivate the firm to build a DC (Lichtenthaler 2009; Menguc and Auh 2006); and frequency of changes to operations—to control for the need to have a DC as a specialized method of changing operating capabilities (Karim 2009; Tzabbar 2009).

A third proportions test revealed that significantly over 50 % of entrepreneurial ventures reported having DCs.

The probit provided a significant improvement in the ‘hit rate’; from 51 to 62 %.

We also checked for the robustness of the results to industry dynamism (not formally reported); we ran the analysis with an interaction term between DC activity and industry turbulence—the key results remained, and this term was not significant.

DC value, rarity, inimitability and non-substitutability were rated on a 5-point Likert scale (from Not at all to Very much). DC costs were rated on a 5-point Likert scale (from Less than 20 % of Gross Benefits to More than 80 % of Gross Benefits).

The process of gaining a competitive advantage—of redeploying resources (in new combinations; Schumpeter 1934) ahead of rivals to exploit an opportunity—is exactly the same process for each of the DCV and ENT ‘views’. The conceptual support of the rents for both views is based on first mover advantages (Lieberman and Montgomery 1988) and on the exploitation of spot arbitrage opportunities (i.e., buying and creating factors under their true future value). The sustainability mechanisms vary, but those of both views include: second-mover disadvantages; market frictions; property rights; and time compression diseconomies (Dierickx and Cool 1989). The usual distinction is that the DCV rents are sustainable in the longer-term than ENT rents. We make two arguments against that distinction: (1) it is a post hoc distinction at best at the empirical level; and (2) such longer-term sustainability is usually described as a series of ‘temporary’ advantages, the type of advantages at the core of ENT theory.

Entrepreneurial ventures will likely differ from larger, older firms in their use of DCs in several ways, including: the timeframe involved (shorter for the SME); the size and scope of the intended impact (smaller for the SME); the OC target type (e.g., technological versus market-oriented); the frequency of use (lower for the SME); the strategic stance (less reactionary for the SME); and the timing (earlier in the industry lifecycle for the SME; e.g., Branzei and Vertinsky 2006). Also, entrepreneurial ventures will likely differ from larger, older firms in the intended and realized outcomes from their use of DCs in several ways, including: the pace of the results (faster for the SME); the sustainability of the results (less for the SME); the variance of the results (higher for the SME; e.g., Ahuja and Lampert 2001); and the rival reaction to the initial results (less reaction from larger rivals for the SME).

The data collection was a one-time survey, although it requested the respondents account for past variable values. This was not true panel data, where unobserved heterogeneity could be controlled; this is a limitation of the paper, and we recommend follow-on work be longitudinal. Despite this limitation, we believe that the results provide a contribution to the literature.

For example, in the DCV, the issue of measurement is problematic. The DC concept is not consistently defined in theory or practice (Williamson 1999; Winter 2003), so it is not surprising that the study of the relationship between DCs and performance involves different methods and controls (Arend and Bromiley 2009). Thus, we advise caution in applying our results (or any DCV study’s results) in a ‘general’ fashion.

There is no limit on the possible variables that could have been included to explain performance, and so omitted variable bias is always a concern in a study like this—i.e., a study on SME performance using survey data. We followed the precedent of related studies by including the main controls—i.e., on firm characteristics, contextual conditions, and entrepreneurial orientations. We were limited on the number of variables we could include by the sample size, by concerns regarding multicollinearity, and by the constraints related to using a survey (e.g., limits due to participant attention). Our regression significances and VIFs were reasonable, however, indicating that we achieved a similar balance in analysis to related studies (e.g., Grant 2012; Simsek et al. 2007). That said, future studies could include controls for other factors, such as those related to the entrepreneur (e.g., social capital), the top management team (e.g., network characteristics), founding conditions, intellectual property, alliance activity, new product development, R&D and advertising intensities, international presence, and other factors that have been specifically studied in strategic entrepreneurship in the past in order to help explain the remaining variance in firm performance.