Who do I hire first?

the effect of first year employees’ social capital on firm survival

Bram Timmermans

Aalborg University

Department of Business Studies

Fibigerstræde 4,

DK-9220 Aalborg Ø, Denmark

Email:

2nd of January 2008

(First draft please do not quote)

1.Introduction

This paper takes the point of departure within labor mobility and how the previous location an individual was situated influences the performance of the current firm. This concept, which is the phenomenon where an employee leaves an employer and starts working in another form of employment (Mincer and Jovanovic 1981 and Elias 1994), related to performance is evident when focusing on newly established firms. Especially since most of the founders of these new ventures were employed before they started their own business (Sørensen and Philips 2004). There exists already a line of literature that links the characteristics of the previous workplace with the performance of the new established firms, e.g. better performing firms will produce better performing ventures (Klepper 2002, Dahl and Reichstein 2006) and smaller firms produce better entrepreneurs (Sørensen and Philips 2004), which make the importance of labor mobility more evident. One of the reasons why the previous workplace is important is related to its function as an organizational blueprint for the new established firm (Baron et al. 1999). Besides the blueprint factor, as a result of experience, the contacts obtained in this firm are regarded as another factor determining the new firm performance (Burton et al. 2002) and it is these contacts that are the main interest of this paper.

Whenever these contacts contribute to a better performance this organization is able to extract benefits from the relations with these contacts. In other words one is intended to speak of this phenomenon as social capital (Portes 1998). However, the contacts the new firm has are not only established at the previous workplace. Typical contacts that are important for newly established venture, but have been established in a more private setting, are the relations with family and friends. Greve and Sallaf (2003) thereby stress that these ties are important, for the founder, in all the phases of the entrepreneurial process. All in all these ties enable access to for a new firm important sources e.g. financial capital, human capital and knowledge capital (Sorenson 2005).

It is thus very logic that organizational researchers are interested in the social capital of founders (Davidsson and Honig 2003, Audretsch and Keilbach 2004,Fornahl 2005, Dahl and Sorenson 2007) but the picture is not yet complete. What should be taken into consideration is that one of the first tasks a founder will undertake is recruit others to join the firm (Dahl and Klepper 2007). Since a large share of these new employees are, just like most of the founders, already experienced and the fact that in the first years the number or new employees will be small the background of these first year employees would influence the performance of the firm using their organizational blueprints and contacts. This will thus not be an attribute only possessed by founders and co-founders. This paper will thus look at the social capital of these first year employees and the effect on the performance, in terms of survival, of the new established firm.

For the empirical angle of this paper I will use the Danish annual census dataset. With this dataset I will identify the potential size of the employees’ social capital and how this subsequently affects the survival of the firm where these employees are employed. The longitudinal character of the data is very suitable for identifying social capital since one of the important requirements for building social capita.

In the next section I will shortly explain the concept of social capital. Thereby choosing the dichotomy that I will use throughout the paper and setting up the hypothesis. In Section 3 I will continue with describing the Danish census dataset, sample selection and the construction of the variables. Section 4 will present the descriptive statistics followed by the results of the logistic regression and I will conclude this paper with a short discussion and concluding remarks in Section 5.

2. Theory and Hypotheses

Social capital has received increasingly more attention ever since Coleman (1988) presented his work on the link between social capital and human capital. The concept itself refers to, as I already mentioned in the introduction, the ability to obtain benefits from the actors to whom another actor is connected (Portes 1998) and although it can be treated both on the individual and organizational level it is primarily an attribute ascribed to individuals (Davidsson and Honig 2003). The social connections of an organization are after all determined by the social connections the members of this organization have. However, the organizational component becomes important later on in this section when I discuss how to treat social capital in this paper.

One of the attributes that can be obtained from one’s social contacts are, first of all, other forms of capital like financial capital, human capital, and knowledge capital (Burt 1992, Sorenson 2005). What has to be taken into account however is that these forms of capital are mostly temporary and borrowed (Lin 1999) but available for use.

Another attribute of social capital is that the relation, or a tie between individuals, is built on trust (Putnam 1995). Building of trust, however, requires two main elements i.e. time and interaction. Thus, for building social capital it is required to know the contact for a certain period of time and during this period have frequent interaction, often accompanied by geographical proximity (Fornahl 2005, Dahl and Sorenson 2007).

There are different ways in which these ties can be categorized. One of the most familiar categorizations is between strong and weak ties as identified in the classic work by Granovetter (1973) where strong ties are those ties with friends and family and weak ties the relation with acquaintances. Strong ties would provide a relatively stable access to resources. Weak ties, however, will provide access that would normally be more difficult to obtain. Because of the loose connection it is information that none of your strong ties would have since your strong ties in a network are often very connected among as well since they meet in similar settings. This is what Burt (1992) identifies as the lack of structural holes (Burt 1992). Another classification, and the one that will be adopted in this paper, is the one mentioned by Adler and Kwon (2002) between bridging and bonding ties. Here the organizational dimension, to what I hinted in the first paragraph of this section, comes forward. The organizational setting in which actors are located determines whether a tie is bridging or bonding.

Whenever an actor has a tie with an individual that is located outside the organization then one is intended to speak of a bridging or external tie. Bridging ties provide access to unique knowledge and contacts (Beckman et al. 2007). Thus, resources that otherwise would not be available for the firm (McEvily and Zaheer 1999). Although this description shows much resemblance to the strength of weak ties argument of Granovetter (1973) a bridging tie does not necessarily have to be a weak tie. Bonding ties are relations that exist inside the organization and build much on the degree of trust and cohesion within the firm (Beckman et al. 2007). There is clearly a trade off between bridging and bonding ties. Whenever a bonding tie exists between individuals in the organization there is high likelihood they know the same bridging tie. For accessing more bridging ties the share of bonding ties in an organization would have to decline. However both ties are considered important for the performance of newly established firm (Davidsson and Honig 2003).

2.1.Bridging Ties Hypotheses

Bridging ties are thus ties an individual possesses but one that exists outside the organizational setting of which this individual is member of. I will start out formulating the hypothesis by looking at the contacts obtained at the previous workplace, as I started this paper by mentioning the importance of the previous firm.

Burton et al. (2002) explicitly mentioned that established relations in the previous firm determine new firm performance. I will make a distinction between two types of relations that could have been established while being in this previous firm, which are (1) relations between colleagues and (2) the relation with actors external to the previous firm. The likelihood of establishing a tie with a colleague is high, especially since the interaction with this group occurs on a daily bases. Thereby is the strength of a tie with a colleague positively correlated with the period a person works for the same firm. Whenever a person leaves the firm the connection with his former colleagues will not disappear over night. In previous studies one finds even support that former colleagues, after current colleagues, are still an important source of information (Schrader 1991, Dahl and Pedersen 2003). Since the value of the social capital decays the longer individual have been apart from each other I assume that the ties are strongest with former colleagues an employee had at the previous firm. During a period of employment a person comes in contact also with individuals and organizations outside the firm with whom they regularly interact e.g. customers, suppliers, supportive industries, etc. These contacts might also be valuable for future professions. Work tenure will, just like the strength of a tie with a former colleague, determine the strength of these ties. The first hypothesis will thus be:

Hypothesis 1: A long tenure at the previous firm increases the likelihood of firm survival.

However, having been outside the labor market for a certain number of years between the previous firm and the current firm means a higher chance of losing contacts. Being outside the labor market can be considered as an element weakening the bridging ties an individual possesses. Thus,

Hypothesis 2: The longer the employee’s period of “between jobs” before entering the newly established firm lowers the likelihood of firm survival.

The interaction with colleagues occurs often because one is closely located to each other, at least regarding the fact that they would work in the same building. Social capital is thus in most cases confined in a certain geographical space and having experience, and thus likelihood of having social ties, in a certain region can be considered as important for having many social connections (Dahl and Sorenson 2007). Dahl and Sorenson (2007) have made a study looking at the location behavior of newly established venture. In their researched they proved that those ventures established in the region where the founder has previous experience have lower failure rates. In the starting up process of a new firm there would be a high dependence on bridging ties. Not only for getting access to attributes like human and financial capital but also from the viewpoint of social support .

The same would also be valid for these first year employees in new established ventures. They might, just as the founder, have essential connections that help the firm survive. However, there is most likely a difference in the relative value of some of the ties both possess. In the case of the founder the social ties will already come on the scene before the business is up and running (Greve and Sallaf 2003). Resources need to be gathered and support by family and friends is a crucial element. There is thus a heavy reliance on social ties regarding professional matters and on those ties with a high private character e.g. with family and friends. Fornahl (2005) groups the motivation of the establishments of these ties in two categories. The first one are ties that are established because the members of a certain network have the same goal, i.e. the professional tie, and there are ties that develop for other reasons, i.e. private ties.

I argue that for employees having experience in the same commuting region is important. With this experience the employee might come more easily in contact with the contacts as they were already identified when discussing Hypothesis 1 i.e. former colleagues and contacts like suppliers, customers, and supportive industries. One might question whether or not the reliance on ties with family and friends is crucial for the survival of the new firm. I argue that this is most likely not the case. Thus although an employee has lived in the current location and thus obtained many bridging ties from a personal perspective these ties will not be as valuable as those bridging ties that have a more professional character. In case of the employees the following two hypotheses will be formulated:

Hypothesis 3: a high share of employees that worked in a different region while they worked for the previous workplace brings a lower likelihood of survival.

And:

Hypothesis 4: a high share of employees that since the previous job moved to a new region is negative for firm survival

For Hypothesis 3 I am interested of the experience in the commuting region. A commuting region consists of several municipalities and the professional contacts might thus be spread of a larger region then the municipality where most likely most of your personal social ties will be residing. Therefore the municipality is the observatory location in Hypothesis 4.

2.2.Bonding Ties Hypotheses

Bonding ties are thus the relations that exist inside the organization, which means, in this case, ties that already existed between members of the organization before starting a career in the new firm. There are different settings in which these previous ties were created e.g. friendship, family, former classmates, former colleagues etc. In this paper I decide to only focus on the ties that exist because the members of the organization shared the same experiences in a previous workplace. In the case of this paper it means that I will look again at the previous firm and see whether or not the employee shared the same previous workplace with respectively the founder(s) and other employees.

Beckman (2006) described, based on already existing literature, some benefits of having shared experiences in the same firm; i.e. shared language, common perspective, degree of trust, common frame, shared vision and set of goals, and a conceptual filter. The tie that exists is most likely freed from conflicts; otherwise they probably would not want to join the same firm. These bonding ties will also not bring resources from outside the organization (Beckman et al. 2007). However, in general the cohesiveness in the firm will be strengthened the more bonding ties exist. Since I make a distinction between the relation between employees and founder(s) and among employees the hypotheses will be:

Hypothesis 5: high degree of previous firm work experience between employees and founders increases the chance of firm survival.

And:

Hypothesis 6: high degree of previous firm work experience among employees increases the chance of firm survival.

3. Data and Method

3.1. The Dataset

For the analyses of the effect of first year employees on the survival of newly established firm I will rely, as many did before me, on the Danish Integrated Database for Labor Market Research which is maintained by Statistics Denmark. When referring to this database I will do so by using its Danish acronym IDA. The information that is contained within this dataset makes it very suitable for this study for the following reasons.

Since the database is built up from government registers it contains information regarding all individuals that (legally) reside in Denmark in a given years. This information has been collected every year, since 1980, and is thus an annual census of the population of Denmark.

For all these individuals IDA has information regarding their characteristics e.g. gender, age, highest fulfilled education, annual income, nationality, etc. Information regarding their labor market status is also provided, which makes it possible to track down in which firms, industries and region this person has worked and what their position within a firm has been. Because the longitudinal character it is possible to follow their career track since 1980. Thereby linking employees and employers to firm and measure characteristics such as tenure, work experience and connections with colleagues.

3.2. The Sample

For constructing a sample I set a number of requirements. Currently I am only interested in those firms that, besides the founder, have other employees and which were founded in 2000. To assure that these firms are newly established I use both the plant and the firm identification number. Each firm may consist of multiple plants and since I am able to identify the year of establishment through the plant identification number I will omit those firms, which have a plant that was established before 2000. However, that all the plants in a firm might have been established in the same year does not automatically mean that the firm did not exist previously. The next step is to remove those firms that have a firm identification number that existed between 1980 and 1999. As mentioned previously a firm might consist of multiple plants. For the analysis in this paper I will only focus on those firms that consisted of only one plant in the year of establishment.