Skip to main content
Log in

Family firms, political connections, and R&D activities in Eastern European Countries

  • Original Paper
  • Published:
Empirica Aims and scope Submit manuscript

Abstract

Although R&D activities can be effective means for firms to develop their innovation capacity, the current understanding of which informal institutions affect firms’ propensity to invest in R&D remains limited, especially in emerging and less-developed countries. Using a large sample of transition countries, this work investigates whether family ownership and political connections influence the firm’s propensity to invest in R&D, as well as in a specific type of open innovation (OI) strategy (i.e. performing simultaneously internal and external R&D activities). According to our evidence, both informal institutions seem encouraging firms to invest in R&D. Moreover, family and political connections appear increasing the engagement in the aforementioned OI strategy. Finally, the two kinds of networks seem to substitute each other in boosting the probability of investing in internal and external R&D activities.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1

Similar content being viewed by others

Notes

  1. Brinkerink and Rondi (2021) show that the family institution can play an “institutional void filling” also for firms located in advanced European countries, making firms more prone to allocate their resources to research and development.

  2. Our analysis complements that thin stream of research that analyzed how institutional voids in transition economies lead firms to adopt a network-based strategic approach, where the growth of firms is promoted by the shared access to tangible and intangible assets (Danis et al. 2009; Michailova and Worm 2003; Wright et al. 2005).

  3. In general, dichotomous choice models can be illustrated in terms of an underlying latent variable process. In our case, we can assume the existence of a latent propensity to invest in R&D, indicated by f*, generated by the following process: \(f_{i}^{*} = X^{\prime}_{i} \beta + u_{i}\), where \(u_{i}\) is an error term and the vector X includes the potential determinants of firms’ R&D. When \(f^{*} > 0\) one observes the phenomenon under study. If \(\delta\) is an indicator function such that \(\delta = 1\) if \(f^{*} > 0\) and \(\delta = 0\) if \(f^{*} \le 0\), the probability of observing firms’ R&D is \(P\left( {\delta_{i} = 1{|}X_{i} } \right) = P\left( {f_{i}^{*} > 0} \right) = P\left( {u_{i} > - X^{\prime}_{i} \beta } \right) = F\left( {X^{\prime}_{i} \beta } \right)\) where F is the standard normal distribution function.

  4. On the other hand, size can adversely affect R&D activities. Situations such as loss of managerial control or excessive bureaucratic power can compromise the activity in R&D (Cohen 2010). Moreover, other studies find a U-shaped relationship where both small and large firms are more productive in R&D than medium-sized firms (Tsai and Wang 2005).

  5. However, older firms could be financially stronger and with more market power and could exploit the ‘learning by doing’ or ‘accumulated learning’ effects by implementing more efficient, cost-effective and routinized processes, thus becoming more R&D intensive (Cohen and Klepper 1996; Mishra 2007; Banerjee and Gupta 2021).

  6. We did not include a measure of profitability because total costs are available only for a limited number of observations. However, as results available on request show, our findings are confirmed when we estimate a model, including costs, on a restricted sample.

  7. Indeed, there is a significant gap in the risk-taking propensity of men and women that is also determined by the culture of the country in which they operate (Mueller 2004).

  8. By contrast, foreign owners can provide access to more progressive technology which may reduce the need of a firm to invest in R&D activities (Un and Cuervo-Cazurra 2008).

  9. On the other hand, state-owned firms may have greater access to credit (Inoue et al. 2013), and obtain additional resources for R&D investments (Lewellyn and Bao 2021). Moreover, state ownership could relieve the pressure stemming from regulatory supervision and provide privileged access to government funds (Boeing et al. 2016), making investments in R&D less risky. According to Zhou et al. (2016) state ownership could support the R&D activities of firms even in an emerging economy.

  10. Family status and political connections could be related to particular types of industries. Yet, ‘empirical evidence seems to suggest that family firms do not decide to self-select out of specific industries but are present in most sectors’ (Berrone et al. 2010, page 111).

  11. Using the same notation of footnote 3, the unobserved latent variables can expressed as follows: \(f_{1i}^{*} = X^{\prime}_{1i} \beta_{1} + u_{i1}\) and \(f_{2i}^{*} = X^{\prime}_{2i} \beta_{2} + u_{i2}\), while the outcomes are \(\delta_{1} = \left\{ {\begin{array}{*{20}l} 1 \hfill & {\quad if\;f_{1i}^{*} > 0} \hfill \\ 0 \hfill & {\quad if\;f_{1i}^{*} \le 0} \hfill \\ \end{array} } \right.\) and \(\delta_{2} = \left\{ {\begin{array}{*{20}l} 1 \hfill & {\quad if\;f_{2i}^{*} > 0} \hfill \\ 0 \hfill & {\quad if\;f_{2i}^{*} \le 0} \hfill \\ \end{array} } \right.\).

  12. At the bottom of column 1 of Table 4, the result of the Wald test indicates that the ρ parameter is highly significant in both estimations, indicating that the error structures of the equations are correlated.

  13. The EBRD-EIB-WB Enterprise Surveys (ES) are a joint initiative of the European Bank for Reconstruction and Development (EBRD), the European Investment Bank (EIB) and the World Bank (WB). They succeed the Business Environment and Enterprise Performance Surveys (BEEPS), which have been carried out in several rounds (1999, 2002, 2005, 2009 and 2012–2014), with the primary goal to provide indicators of the business environment and firm-state interaction, in Central and Eastern Europe and the former Soviet Union. The ES survey includes 41 countries (of EU, Eastern Europe, Central Asia and Middle East and North Africa). In this analysis, we consider the following countries: Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, the Czech Republic, Estonia, Georgia, Hungary, Kazakhstan, Kosovo, the Kyrgyz Republic, Latvia, Lithuania, Moldova, Mongolia, Montenegro, North Macedonia, Poland, Romania, Russia, Serbia, the Slovak Republic, Slovenia, Tajikistan, Ukraine and Uzbekistan.

  14. According to a strand of literature, the family institution could compensate for the weakness of other formal and informal institutions, exerting a particularly positive influence on the performance of firms in countries where institutions are less developed (Miller et al. 2009; Mengoli et al. 2020). Analogously, political connections are likely more pronounced ‘where markets fail and institutions are weak as characterized by weak rules of law, rampant corruption, government control of the press, a lack of accountability and transparency, government intervention in business activities and low-quality public governance’ (Ha and Frömmel 2020, page 2).

  15. Similar to Charron et al. (2019), we compute a composite index, as the country average of five indicators: ‘control of corruption’, ‘government effectiveness’, ‘rule of law’, ‘voice and accountability’ and ‘regulatory quality’. Moreover, as a further robustness check, we also employ measures of Regulatory Efficiency, Rule of Law, Open Market and Government Size that are the main pillars of the Economic Freedom Index (Heritage Foundation). We also employ the International Property Rights Index (Property Rights Alliance) that is based on three dimensions: physical property rights, intellectual property rights, and the legal and political environment. The relative output is available upon request.

  16. Our external IVs are: the average values of FAMILY and POLITIC computed by sector and country, which are significantly correlated with the two potentially endogenous dummies, after controlling for the other explanatory variables. Indeed, firms operating in a sector within a country are likely to share some characteristics, some of which can influence the diffusion of family ownership and political connections. At the same time, the diffusion of family control and political networks within a sector in a country should not be directly correlated with the R&D activities of a firm. According to the Kleibergen-Papp LM test and the Hansen j-test statistic (bottom of Table 3), the null hypothesis that the model is under-identified is rejected, whilst the null hypothesis that the instruments are valid cannot be confuted.

  17. The results for the control variables are quite similar to the results obtained by applying the Probit model.

  18. Looking at the magnitude of the estimated effects, the probability of doing internal R&D increases by 2% for family firms with respect to the non-family counterparts. Analogously, the probability of acquiring external knowledge increases by 2% for politically connected firms. Finally, the probability of conducting both activities simultaneously increases by 2% for family firms with respect to the non-family counterparts and by 3% for politically connected firms.

  19. The above-mentioned complexity stems from the fact that in non-linear models the marginal effect of the interaction does not coincide with the first derivative of the multiplicative terms. Indeed, the full interaction effect is the cross-partial derivative of the expected value of the outcome variable E(y) with respect to the constitutive terms (for instance, \(x_{1}\) and \(x_{2}\)), which is different from the first derivative of E(y) with respect to the multiplicative term (\(x_{1}\)*\(x_{2}\)). Besides, the statistical significance of the entire cross-derivative must be calculated. Finally, like all marginal effects in non-linear models, the interaction effect is conditional on the other independent variables, and it can have different signs for different values of covariates.

  20. The “Appendix” reports further graphs based on the Bivariate Probit estimation, corroborating the finding that family businesses seem to privilege in-house R&D, whilst politically connected businesses appear more capable of exploiting external sources of knowledge.

References

Download references

Acknowledgements

We are grateful to an anonymous reviewer and the participants of the SIEPI and SIE 2022 conferences for their precious comments and suggestions. We also wish to thank the Editor in Chief Harald Oberhofer. Any remaining errors are solely our responsibility.

Funding

No funding was received for conducting this study.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Cristiana Donati.

Ethics declarations

Conflict of interest

The authors have no potential conflicts of interest to declare.

Research involving human participants and/or animals

This research does not involve Human Participants and/or Animals.

Additional information

Responsible Editor: Harald Oberhofer.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendix

Appendix

Based on bivariate Probit estimation, the following graphs show the fitted probability of two decisions: performing only internal R&D (\(R\& D_{IN} = 1\) and \(R\& D_{OUT} = 0\)) or outsourcing (\(R\& D_{IN} = 0\) and \(R\& D_{OUT} = 1\)), for family (red) and non-family firms (blue). The x-axis displays the values of the dummy variable POLITIC. In Fig. 

Fig. 2
figure 2

Fitted probabilities of investing only in internal R&D (with 95% cofidence interval)

2, when focusing on firms non-politically connected (POLITIC = 0), the estimated probability of conducting in-house R&D is significantly higher for family firms with respect to non-family firms. On the contrary, the ownership of the firm seems no longer relevant for politically connected businesses, as the confidence intervals overlap when POLITIC = 1. Moreover, political connections do not significantly influence the probability of carrying out internal R&D either for family or for nonfamily firms, as both the blue and the red intervals on the left overlap with the analougous intervals on the right.

Looking at Fig. 

Fig. 3
figure 3

Fitted probabilities of investing only in external R&D (with 95% cofidence interval)

3, the fitted probability of acquiring knowledge from external sources \(\left( {R\& D\_IN = 0 and R\& D\_OUT = 1} \right)\) is not affected by family ownership either when firms are politically connected or not, as the two confidence intervals overlap either when POLITIC is 0 or 1.

Moreover, being politically-connected appears to increase the probability under investigation both for family and non-family firms, as neither the blue nor the red interval on the left overlaps with the analougous interval on the right.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Agostino, M., Donati, C. & Ruberto, S. Family firms, political connections, and R&D activities in Eastern European Countries. Empirica 50, 725–754 (2023). https://doi.org/10.1007/s10663-023-09568-x

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10663-023-09568-x

Keywords

JEL Classification

Navigation