Successfully defended on the 24th of June 2010
This thesis, with the title: ”Human Resources and Firm Performance: A Systemic Perspective on the Human Resource Composition of New and Established Firms”, takes its point of departure in the widely held belief that human resources are an important factor in explaining firm performance. Despite this belief, many questions have remained unanswered in how, and in what way, these human resources contribute to the success of firm, especially from a larger and more systemic perspective. The access to unique and longitudinal linked employer-employee databases offer the means to answer many of these questions and provide insights in how the human resource composition of firms influence firm performance (a digital copy is available upon request).
The research objective of this thesis is, as I was granted access to the Danish linked employer-employee database, to provide this insight. By using this database, better known under the name the Danish Integrated Database for Labor Market Research, I tested the impact of human resources on firm performance by looking at labor mobility, social networks, and organizational demography. Because these different perspectives are presented in the separate article-based chapters, I will give a summary of each individual chapter.
Chapter 1, the synopsis of this thesis, presents a reflection on the findings of all the different chapters in relation to the above-mentioned research objective. In order to do so, the synopsis starts with a description on the importance of human resources in explaining firm performance by referring to human capital theory, the resource based view of the firm, and social psychology theories. Afterwards, I start a discussion on organizational demography, which is the method used to measure differences in the human resource composition of firms. The synopsis will conclude by presenting an overarching discussion on how the impact of this human resource composition differs in different stages of the organizational life cycle. This discussion is initiated by the fact that the different chapters show that young firms benefit from similarity while established firms need to have a diverse human resource composition to outperform others. An explanation on why this is the case might be that young firms face high levels of uncertainty in getting their business established. Similarity in their human resource composition creates a social environment that copes better with this uncertainty. Established firms reduced their uncertainty on many of these variables but they run the risk of organizational lock-in. A diverse human resource composition might assist in breaking through this organizational lock-in. These two perspectives have been extensively discussed by organizational ecologist in their reference to the various age related liabilities, i.e. liability of newness, liability of adolescence, liability of obsolescence, and liability of senescence. This synopsis will discuss these age dependent liabilities further.
The purpose of Chapter 2 is to give a detailed description on the structure and content of the above-mentioned Danish Integrated Database for Labor Market Research (IDA). Despite the broad international interest of this database and its use in a wide range of scientific disciplines, varying from labor market economics to health science, no detailed English documentation on the database existed. IDA provides detailed information on all individuals and all establishments from 1980 onwards. The database is characterized of having a vertical and a horizontal dimension. With a horizontal dimension is meant that individuals, establishments, and firms, due to their unique identification number, can be followed over time. The vertical dimension refers to the division of IDA in three types of databases, i.e. personal, employment, and establishment level data that can be merged together on the desired level of aggregation, e.g. establishment, firm, industry, and/or geographic location. In addition to describing the structure and content of IDA, I also discuss the procedure of selecting start-ups and founders, and suggest an alteration in the procedure of selecting Danish labor market regions. This is discussed in more detail because these entities are a crucial component in the various analyses.
In Chapter 3, which is based on a co-authored paper with Ron Boschma, we test the impact of different types of labor mobility on the performance of establishments. This study takes its point of departure in the often-assumed benefits of labor mobility as a mechanism by which knowledge gets distributed between firms. What these studies do not take into account are the types of knowledge and skills that are transferred between firms through job-hopping. Recently, studies have shown the importance of relatedness in knowledge ands skills, i.e. the need of some degree but not too much cognitive proximity, is important to enhance interactive learning and the performance of firms. We therefore argue, in concurrence with earlier studies, that the effect of labor mobility can only be assessed when one accounts for the type of skills that flow into the establishment, and the degree to which these match the existing set of skills at the establishment level. To investigate this claim for Denmark, we selected those establishments that during the period 1999-2003 experienced an inflow of at least one high skilled worker, i.e. an individual that either is in possession of an academic degree of belongs to the top 20 percent wage earners (n=22,788). Based on the past industry experience of these high skilled workers, we measure the inflow of similar, related, and unrelated skills by comparing this industry experience with the industry in which the establishment is active. Furthermore, we also identify whether this inflow is intra- or inter-regional. As expected, we found that the inflow of skills that are related to the set of skills in the establishment has a positive impact on the performance of the establishment, while inflows of skills that are similar or unrelated to the existing set of skills in the establishment have a negative effect. In addition, intra-regional skilled labor mobility had a negative effect on performance more in general, while the effect of inter-regional labor mobility depends on the type of skills that flow into the establishment. In these analyses, we used a sophisticated indicator of revealed relatedness that measures the degree of skill relatedness between each pair of sectors on the basis of the intensity of labor flows between sectors. We made the same estimations using the more common NACE-based skill relatedness indicator. Although our main findings remained the same, we found that our revealed relatedness indicator generated stronger levels of significance.
Chapter 4 takes a social network perspective toward the human resource composition of start-ups. New firms often fail because they operate in an environment in which they lack the personal relationship and inter-organizational linkages that are necessary to survive, i.e. they suffer from liabilities of newness/adolescence. In this chapter, I investigate the impact of involving former colleagues into the new venture. I argue that start-ups can overcome these liabilities by recruiting former co-workers because: (i) their presence indicates a level of trust between among the workers, (ii) their shared work experience brings contributes to the creation of an organizational culture, and (iii) the former workplace functions as an identifier of the skills that are necessary for the start-up to survive. The overall hypothesis on the positive impact of previous co-worker experience is tested on a sample of Danish start-ups founded in the year 2000 that are all active in private, non-agricultural industries and employ at least two individuals (n= 3,034). To identify their survival rate, these start-ups are followed up to the year 2005, which leads to a total of 10,540 yearly observations. Moreover, I also make a distinction between founders and employees, and whether the start-up can be identified as an entrepreneurial spin-off, i.e. one of the founders has experience in the same four-digit NACE industry class. To measure previous co-worker experience, I selected the three most recent establishments in which each individual was active before joining this new venture. Overall, the different analyses present, after correcting for the usual predictors of firm survival, e.g. human capital, industry, and size, a significant and positive effect of previous co-worker experience on firm survival. Interestingly, this co-worker experience is only positive in relation to the founder of the start-up. Furthermore, previous co-worker experience does not have a significant impact in those start-ups that can rely on the same industry experience of the founder(s), i.e. entrepreneurial spin-off. Extreme levels of co-worker experience are also recommended to be avoided, as there are signs that “too much of a good thing” hampers the likelihood of firm survival.
Chapter 5, which is co-authored with Christian Østergaard and Kari Kristinsson, investigates the relation between the human resource composition of firms and their likelihood to innovation. The link that is made in this chapter is that a diverse composition in a firm’s human resources contributes positively to the likelihood of introducing an innovation, because this employee diversity might create a broader search space and make the firm more open toward new ideas and more creative. The impact of a diverse employee composition on the likelihood to innovate is tested on a sample of nearly 1,700 firms that participated in the DISKO 4 survey on technological and organizational change. The answers to the questionnaire are merged with IDA, which allow us to identify whether a firm introduced an innovation and the level of diversity in their employee composition. Diversity is in this chapter treated as a unit-level compositional construct based on four demographic characteristics, i.e. age, gender, ethnicity, and education. The logistic regression analysis reveals, whenever using size and industry fixed effects, a positive relation between diversity in education and gender on the likelihood of introducing an innovation. Furthermore, we find a negative effect of age diversity and no significant effect of ethnicity on the firm’s likelihood to innovate. In addition, the logistic regression reveals a positive relationship between an open culture toward diversity and innovative performance. We find no support of any curvilinear relation between diversity and innovation.
The goal of Chapter 6 is to describe the procedure of selecting a sample of knowledge intensive start-ups that is being used in the analysis in Chapter 7. A common systemic approach to select these types of start-ups is to select all the start-ups that are active in the so-called technology and knowledge intensive industries. However, given the discrepancy between individual firms and the aggregated industry level, i.e. firms in knowledge intensive industries are not necessarily knowledge intensive, and the presence of knowledge intensive start-ups in traditional non-knowledge and non-technology intensive industries I propose a different approach, especially given the detail information that is available in IDA. Instead of relying on this industry classification, I select knowledge intensive start-ups based on the human resource composition of start-ups in general. In this case, I adopt a person centric perspective toward knowledge intensity. This approach is common in studies that investigate knowledge intensive firms. To requirements to separate knowledge intensive start-ups from non-knowledge intensive start-ups is the presence of at least two individuals that are identified skilled, i.e. either having an academic degree of belong to the top 20 percent wage earners, and the number of skilled workers should consist out of at least 25 percent of the entire workforce in the start-up. Based on these criteria, I identify a total of 6,815 knowledge intensive start-ups in the period 1995-2004. Almost 50 percent of these start-ups are active in the technology and knowledge intensive industries as defined by OECD and EUROSTAT. In addition to identifying these types of start-up, I also present some overall descriptive statistics, e.g. industries in which they are active, their size, their human resource composition, and their geographic location.
In the last chapter, Chapter 7, I investigate the role of demographic diversity in knowledge intensive start-ups. In the last two decades, there has been an increasing interest in the role of entrepreneurial teams and how the composition of these teams influences their performance. In this chapter, which takes a human capital and a social psychological perspective, I will explore in what way this composition affects the likelihood of survival. Due to the proven ”mixed blessing”of diversity, I will keep an open mind on its possible impact. However, the existing literature leans toward the negative side on the impact of diversity on the readily observable characteristics and a more positive attitude toward diversity in the human capital attributes. The test the impact of a diverse human resource composition on the likelihood of firm survival I take the sample of 6,815 knowledge intensive start-ups that have been identified in Chapter 6. In a similar fashion as described in Chapter 5, I look at demographic diversity, based on the ascribed (e.g. gender, age, nationality) and achieved (e.g. education and work experience) characteristics, of all the individuals who are present in the first and second year after firm founding. For diversity in education and work experience I make a distinction between the related and unrelated variety measures as presented in Chapter 3. The results of the conditional fixed effects analyses show a predominant neutral and negative effect of diversity on new firm survival, especially in those situations where there is a high degree of diversity on industry experience and educational background. For a start-up it is thus important to avoid diversity as much as possible.