International Choice Modelling Conference, International Choice Modelling Conference 2017

Font Size: 
Like a rolling stone? Heterogeneity in location preferences of early-stage technology based start-ups
Frank van Rijnsoever, Fenna Cerutti

Last modified: 28 March 2017

Abstract


Introduction

Policies to attract early-stage technology based start-ups from abroad are based on the idea that these individuals are like cosmopolitan ‘rolling stones’. Armed with their laptop they are seen to travel from city to city, in search of attractive creative places (Florida, 2004) to live and to do business (Siegele, 2014). However, this somewhat romantic view seems to contradict empirical studies that show most entrepreneurs are likely to start a business in their home region, and move very little (Dahl & Sorenson, 2009; Michelacci & Silva, 2007; Parwada, 2008; Stam, 2009). Entrepreneurs place more emphasis on behavioral attributes like social ties—i.e. being close to family and friends—than on regional economic attributes that can influence the performance of their business (Dahl & Sorenson, 2009; Michelacci & Silva, 2007; Stam, 2007). This apparent contrast suggests that entrepreneurs differ in their mobility preferences. In other words, there is heterogeneity among entrepreneurs. To understand this heterogeneity, this paper aims to understand how different latent classes of early stage entrepreneurs value different  attributes of a region in the choice to locate their start-up. Our analysis reveals latent classes that are consistent with contemporary theories about entrepreneurship and location choices.

 

Attributes

To this end, we identify latent classes through a choice experiment (DCE). Until now, DCEs are commonly used to understand consumer preferences. We derive the for attributes our DCE from the literature on industrial (Hayter, 1997) and entrepreneurial location choices (Ferreira, Fernandes, Raposo, Thurik, & Faria, 2015). This literature makes a distinction between neoclassical, institutional and behavioral attributes. In addition, we conducted interviews entrepreneurs and other stakeholders to verify our attributes. This resulted in the list of attributes presented in Figure 1.

Figure 1: attributes

Methods

We conducted a discrete choice experiment (DCE) on a representative sample of 935 entrepreneurs with early stage technology based start-ups from Western Europe and North America. Respondents received a series of 8 choice tasks (Figure 2 for an example).

Figure 2: Example choice task.

To be able to interpret the latent classes later on, we asked entrepreneurs to report a series of characteristics of their start-up and themselves. We analysed our data in two steps. First, we fitted a latent class model on the DCE data using the Latent Gold program. We explored solutions of between one and five latent classes using different model specifications. Moreover, we explored whether scale classes could be identified. Respondents exhibit different degrees of consistency when making their choices. This consistency is based on the variance in responses (Magidson & Vermunt, 2007). Not taking into account scale can lead to a bias in model estimates (Sælensminde, 2001; Swait and Louviere, 1993). Second, we fitted a multinomial logit model to predict latent class membership, using the covariates.

Results

Our analyses reveal that three classes of entrepreneurs with regard to their preferences for start-up locations. “Region bound kings”, who do not wish to move, “Creative class networkers”, who focus on a good quality of living, and “Serial rolling stones ” who focus mostly on regional economic attributes. Region bound kings fit well with findings that entrepreneurs place more emphasis on personal than on regional neoclassical economic attributes that can influence the performance of their business (Dahl & Sorenson, 2009; Michelacci & Silva, 2007; Stam, 2007). We find that their businesses appear mature relative to their age. This could be a consequence of the lack of intention to move and that these entrepreneurs have invested in the development of their business at their present location instead (for example by hiring employees). Creative class networkers and Serial rolling stones fit more with the idea of cosmopolitan moving entrepreneur (Siegele, 2014) that are the target of policy makers. Yet, both groups have distinct preferences. Creative class networkers seem attracted by a quality of living and still want to remain in reasonable distance of their loved ones. Serial rolling stones don’t care for these attributes, they are primarily attracted by favorable economic conditions, but they are less clear about what they exactly need. The literature suggests that this is because early stage entrepreneurs are unaware of their resource needs (Bruneel, Ratinho, Clarysse, & Groen, 2012; Van Weele, van Rijnsoever, & Nauta, n.d.). Hence, it seems that this class keeps its options open.

Conclusions

Our study shows that taking into account both observed and unobserved heterogeneity is required to understand the location choices of early stage entrepreneurs. The resulting latent classes fit earlier theoretical perspectives on location choice. Unobserved heterogeneity thereby reconciles the apparent contradiction between the ‘rolling stones’ image of entrepreneurs, and the homebound entrepreneur. This evidences that DCEs can be used to better understand apparent contradictions in literatures from different fields.

References

Bruneel, J., Ratinho, T., Clarysse, B., & Groen, A. 2012. The Evolution of Business Incubators: Comparing demand and supply of business incubation services across different incubator generations. Technovation, 32: 110–121.

Dahl, M. S., & Sorenson, O. 2009. The embedded entrepreneur. European Management Review, 6(3): 172–181.

Ferreira, J. J. M., Fernandes, C. I., Raposo, M. L., Thurik, R., & Faria, J. R. 2015. Entrepreneur location decisions across industries. International Entrepreneurship and Management Journal, 1–22.

Florida, R. 2004. The rise of the creative class and how it’s transforming work, leisure, community and everyday life (Paperback Ed.). GEN, New York: Basic Books.

Hayter, R. 1997. The dynamics of industrial location: the factory, the firm, and the production system. BOOK, Wiley Chichester.

Magidson, J., & Vermunt, J. K. 2007. Removing the scale factor confound in multinomial logit choice models to obtain better estimates of preference. Sawtooth Software Conference: 139. CONF, .

Michelacci, C., & Silva, O. 2007. Why so many local entrepreneurs? The Review of Economics and Statistics, 89(4): 615–633.

Sælensminde, K. 2001. Inconsistent choices in Stated Choice data. Transportation, 28: 269–296.

Siegele, L. 2014. A Cambrian Moment: Special Issue on tech startups. The Economist, (January 18th).

Stam, E. 2007. Why Butterflies Don‘t Leave: Locational Behavior of Entrepreneurial Firms. Economic Geography, 83(1): 27–50.

Stam, E. 2009. Entrepreneurship, Evolution and Geography. Papers on Economics and Evolution, No. 0907: 1–22.

Swait, J., & Louviere, J. 1993. The role of the scale parameter in the estimation and comparison of multinomial logit models. Journal of Marketing Research, 305–314.

Van Weele, M. A., van Rijnsoever, F. J., & Nauta, F. n.d. You Can’t Always Get What You Want: How Entrepreneur’s Perceived Resource Needs Affect the Incubator’s Assertiveness. Technovation.


Conference registration is required in order to view papers.