The Impact of the External Environment on the Growth of the Italian Academic Spin-Offs: A Cross-Sectional Analysis

In the last decade, there has been growing attention of the institutions towards the third mission of Universities. The university, along with their two fundamental goals of education and research, pursues a 'third mission', which means that it works to encourage the direct application, enhancement and use of learning to contribute to society's social, cultural and economic development. An instrument to pursue this mission is the creation by universities of academic spin-offs (ASOs) transferring the research results to the business world. This paper tested how the external environment would affect or not the performances of ASOs and in what terms. The dependent variable that best expresses growth, according to the academia, is the sales and the elements of the external environment. It wanted to test if the presence or not, in the province where the ASOs is located, of technology parks, business accelerators and incubators are influential. In this paper, it conducted a cross sectional analysis using an OLS examining the financial statements (31/12/2014) of 552 Italian ASOs.


Introduction
As well as education and research, the University pursues a third fundamental objective (so called third mission), the work that is to promote the direct application, enhancement and use of knowledge to contribute to the social, cultural and economic of the society.In this perspective, any structure within the University is committed to communicate and disseminate the knowledge through a direct relationship with the land and with all its stakeholders.Third Mission activity is a vitally important component of any university's role, whether it pictured as a third mission or as integral to the core missions of education/teaching/learning and research/scholarship.It is as important for the university in countless ways as it is for society.It is not new, but narrower notions of research excellence have overshadowed it, and academics have usually drawn themselves into something of a caste apart.However, Third Mission over the last decade revived.(Green Paper, European Commission, 2008).In Italy, the Legislative Decree n. 19/2012 set out the criteria of the Self-Assessment System, Periodic Evaluation and Accreditation, and then the Ministerial Decree n. 47/2013 has identified indicators and periodic evaluation parameters of research and third mission.They recognized in effect the third mission as an institutional mission of universities, alongside the traditional teaching and research missions (Anvur).Scientific and technological inventions produced by universities and moved to the market by new firms represent an option increasingly used for wealth creation that departs from the transfer of knowledge generated from research (Carayannis et al., 1998;Siegel et al., 2003;Vohora et al., 2004;Clarysse et al., 2005;Lockett et al., 2005).Moreover, the transfer of university-industry knowledge has become one of the main factors underpinning European innovation policies (Mueller, 2006).Different countries have thus created support units aimed specifically at strengthening the links between university and industry, with a strong focus on boosting and facilitating the creation of university-based companies (Criaco et al., 2013).The entrepreneurs, during their activities as students, professors or researchers at a university, acquire technological knowledge or develop a new technology that will, in the future, be used with the support of the university´s business incubator (or another mechanism) to develop a product or a business concept that will be explored commercially by a new venture (Borges & Filion, 2013).Academic spin-offs can be defined in two different sense.In a restrictive sense, as firms established on the basis of intellectual property generated within universities, in which the public body of research is directly present with share capital.In a general sense, where academic spin-offs firms are set up on the basis the subjec Piccaluga in which

Compr
In this ch mentione formulate

Metho
The chos regression data in or used.In t the squar regression change in regressor we use a following Where:

In this w environm
Where: • Ln • β 0 • LF is a dummy variable that represents the legal form of the ASO (1 if is S.r.l. and 0 if is S.p.a.); • STPC is a dummy variable that represents the presence (1) or not (0), in the province of the legal address of the ASO, of a Scientific/Technological Park or Center (Note 2); • BA is a dummy variable that represents the presence (1) or not (0), in the province of the legal address of the ASO, of a Business Accelerator 4 ; • IN is a dummy variable that represents the presence (1) or not (0), in the province of the legal address of the ASO, of an Incubator 4 ; It uses the natural logarithm of total sales (LnSALES) as dependent variable as Wennberg et al. (2011) that estimated a model for growth in terms of sales using the formula log (size 1 /size 0 ) to compute the growth rate.In this paper, using a cross-sectional analysis it estimated a model for growth using the formula log (size 2014 ) to obtain as result the effect, in percentage, on Y of a change in independent variable, holding all the other variables constant.The percentage change in Y is equal to (100 * ∆Y/Y); a change in X by one unit (∆X = 1) is associated with a 100 * β1 % change in Y (Stock & Watson, 2005).The logarithm also has the advantage to transform the sales value in a smaller number easier to use.This is the case of a "log-linear" function (Note 3) because just the dependent variable (and none of independent variables) is a natural logarithm.
The control variables used, pursuant to the most relevant literature (Rodriguez et al., 2016), in this survey are the following: 1. Ln of Age (LnAge) 2. Ln of Total assets (LnAS); 3. EBITDA margin (EM);

Legal form (LF).
The external environment variables (STPC, BA and IN) are the three variables that It analyses with the aim to understand if, and to what extent, there is a relationship (positive or negative but especially statistically significant) able to influence the ASO's Sales.

Analysis
The main findings of the investigation about the relationship between external environment and ASO's Sales presented in this section.The descriptive statistics represented in Table 2.It is important to point out, as shown in Table 1, that the average value of the ASO's Sales is 4,86 (with a median of 4,80).This result is consistent with the one obtained by Lipton & Lorsch (1992).Table 2 illustrates the correlations among different explanatory control and dependent variables used in the regression analysis.Outliers do not affect the dataset because, as it is possible to note in Table 1, mean and median are very similar values.As shown in Table 2, an interesting finding from the correlation matrix is that the presence in the province of the legal address of the ASO of an Incubator is positively related to the natural logarithm of Sales (0,12).Contemplating probably the fact that most of the firms with localized "near" an Incubator could be able to achieve better financial performance than others could.The higher correlation coefficient regards the relationship between the natural logarithm of sales (dependent variable) and the natural logarithm of total assets (0,79), demonstrating that the firm size in a determinant factor to improve the financial performance.
There are not too high correlation coefficients and this could indicate that collinearity does not affect the sample.
The OLS regression results obtained presented in Table 3.The results are robust to the effect of multicollinearity, outliers and non-linearity.It runs a "White Test" to understand if there is a heteroskedasticity problem, (the error terms, do not have constant variance) and the results are that heteroskedasticity does not affect the sample.As is possible to note in Table 3 the control variables (LnAge, LnAS and EM) are statistically significant at 1% or 5% level but in this work we comment and discuss just significant external environment variables (IN).The findings, with regard to the link between the presence in the province of the legal address of the ASO of an Incubator and Sales seem to demonstrate that a positive relationship exists and that this link is statistically significant.Better financial performance, in terms of Sales, characterized firms localized "near" an Incubator.The adjusted R2 value is high (0,636) and it means that more than 60% of the variance in the LnSA explained by the independent variables used in this model.
The highest variance inflation factor (VIF) in the regression model is considerably within the limit as none of VIF approached the critical value of 10, as shown in Table 4.The highest coefficient regards the natural logarithm of total assets and is equal to 1.59.This is the reason why it is possible to assert that multicollinearity (when one of the regressors is an exact linear function of the other regressors) does not affect this dataset.According the aforementioned regression results it is possible write the equation as following, substituting the coefficients of regression in the formula: = −0,04 + 0,13 2014 + 0,75 2014 − 0,015 + 0,06 2014 − 0,07 ℎ − 0,04 + 0,15

Conclusions and Limitations
In conclusion, we can say that the external environment influences the growth of ASOs and especially the presence of incubators is a fruitful context for growth.The presence in the province of the legal address of the ASO of an Incubator involve higher Sales.This statement also supported by what expressed by some of the academic spin-off members interviewed in the course of the present research.They confirmed that proximity to the university or business incubators would allow them to participate in organizing events and take advantage of destructured resource (stage, student internship, ...) that somewhat enrich the spin-offs.
As far as the limits of this research, focusing exclusively on single year data, it is not possible to test the effect of the years on the results themselves and therefore, for future studies, we may want to consider using panel data.Furthermore, in order to clean up the results from reverse causality.We could evaluate the possibility of using a dynamic panel (GMM) or search an instrumental variable with which to integrate the model.

iew of the Th
EM is the EBITDA (Note 1) margin, obtained dividing EBITDA by total sales; this result helps show how much operating expenses are eating into a company's profits.

Table 5 .
Multicollinearity test and VIF