Determinants of investing in innovative activities by agri-food and KIBS firms in rural areas: An exploratory analysis

ABSTRACT

Using a database on innovation activities by Spanish firms, we perform an exploratory analysis on a frequently overlooked area of research: innovation investment decisions by small firms in rural areas where socioeconomic indicators are weak. We focus on two quite different sectors of interest there, agri-food and t-KIBS firms, to explore how the regional context influences the capacity and the nature of innovation. Following the TOE approach as a conceptual framework, we perform a multivariate statistical analysis based on MCA and PCA combined, to identify the most relevant factors among a list of 73 indicators in four broad domains, including innovation decisions as well as organizational, technological, and environmental determinants. The exploratory results obtained suggest an open field of research. Thus, we observe a distinctive behavior of marketing innovation by agri-food firms that is related to former ICT experience, whereas services innovation by t-KIBS would be related to the objective of entering new market niches. Some results confirm previous results for SMEs outside rural areas, such as financial constraints, lack of qualified personnel, and strong competition being relevant barriers to invest in R&D by t-KIBS firms. Contrariwise, we dispute the assertion that agri-food firms are primarily oriented toward product and process innovation: what comes together is product and organizational innovation.

Notes

¹ Five years is fair enough for a statistical analysis with pooled data, while a larger period might be susceptible a structural changes in the decisions to innovate by firms (e.g., before and after the financial crisis). Nonetheless, if we wanted to enlarge the sample, we were limited by the data provided in the PITEC database—in particular, a large set of variables on organization and marketing innovation, and the objectives of innovation, among others, were included only after 2008.

² DEGURBA (http://ec.europa.eu/eurostat/ramon/miscellaneous/index.cfm?TargetUrl=DSP_DEGURBA).

³ All sources of data, Instituto Nacional de Estadística (INE), www.ine.es, except fertility rate in Europe (source: Eurostat).

⁴ Revenues are one of the anonymized variables we use from the PITEC database, together with total investment in physical goods, total investment in innovation, the number of employees in the firm, and in R&D activities in particular. According to the INE, the use of these data for statistical purposes is correct for most statistical and econometric research, since the techniques employed for anonymization—basically, ranking and grouping observations to estimate their averages—leads to “small biases in any case” (source: PITEC methodological note by the INE, available at http://www.ine.es/prodyser/microdatos/metodologia_pitec.pdf; last access: September 2018). Moreover, since the statistical treatment we perform consists of transforming any numerical (anonymised) variables into quartiles, any biases are avoided.

⁵ Any statistical treatment implies loss of information, and this implies some limitations to our research. In particular, the reduction of variables into domains performed by MCA implies an initial loss of information, although inevitable, as there is no univocal relationship between variables. Then, PCA uses MCA results to describe the structure of the relationship between the different domains on which we base our conclusions. However, these are only descriptive observations that do not allow to test theoretical hypotheses. These should first be defined (here, exploratory analysis can be useful) and then tested using confirmatory techniques.

⁶ For such purpose, we use the package psych of the R statistical suite (Revelle 2017).

Additional information

Funding

This research has received the funding of the Program for the Consolidation and Structuring of Competitive Research Units - Research Networks (Redes de Investigación) (Ref. ED341D R2016/014), Proxectos Plan Galego IDT, from the Xunta de Galicia (Spain).