Business start-ups, closures and economic churn
A review of the literature
Final report
Prepared for the Small Business Service
rd23 August 2006
1Catherine Robinson, Brigid O‟Leary, Ana Rincon
Consultants: Mary O‟Mahony and Geoff Mason
URN 06/2112
1 We would like to thank John Forth for helpful comments on earlier drafts. Our thanks also extend to
our colleagues at the SBS, in particular, Daniel van der Schans and Jonathan Gershlick.
Table of Contents Business start-ups, closures and economic churn ..............................................i
A review of the literature .......................................................................................i
Executive Summary ............................................................................................ iv
1. Introduction .................................................................................................. 1
2. The theoretical foundations ......................................................................... 4
3. The definition of economic churn .............................................................. 10
3.1 Unit for consideration – plant or firm? ................................................ 10
3.2 Types of entry and exit ........................................................................ 11
3.3 The importance of industry ................................................................. 13
3.4 The importance of time ....................................................................... 13
3.5 Conclusions ......................................................................................... 14
4. Measurement of entry, exit and churn ...................................................... 15
4.1 Conclusions ......................................................................................... 18
5. Rates of churn ........................................................................................... 20
5.1 Churn at the regional level .................................................................. 23
5.2 Conclusions ......................................................................................... 24
6. The impact of churn on the economy ....................................................... 25
6.1 Productivity .......................................................................................... 25
6.2 Empirical studies of the productivity impact of churn ........................ 27
6.3 Employment......................................................................................... 32
6.4 Economic growth ................................................................................. 34
7. The costs of churn ..................................................................................... 36
7.1 Conclusions ......................................................................................... 39
8. Alternative modes of entry and exit .......................................................... 40
8.1 The entry decision ............................................................................... 40
8.2 Entry by diversifying firms ................................................................... 42
8.3 Changes in ownership......................................................................... 46
8.4 Modes of exit ....................................................................................... 50
8.5 Conclusions ......................................................................................... 50
9. Start-ups and closures: the UK experience .............................................. 52
9.1 Competing theories of the firm............................................................ 54
9.2 Business start-ups ............................................................................... 55
9.2.1 The case for entrepreneurship ........................................................ 58
9.2.2 Industry conditions ........................................................................... 63
9.2.3 Concentration ................................................................................... 64
9.2.4 Barriers to entry ................................................................................ 65
9.2.5 The regional dimension.................................................................... 66
ii
9.3 Technology and innovation ................................................................. 68
9.4 Survival versus exit ............................................................................. 72
9.4.1 The role of size in survival ............................................................... 73
9.5 UK industrial start-up policy ................................................................ 74
9.6 Conclusion ........................................................................................... 78
10. Analysing churn from the top-down ...................................................... 79
11. Proposals for future analysis ................................................................. 81
11.1 Rates of churn ................................................................................... 81
11.2 Churn and productivity ...................................................................... 82
References ........................................................................................................ 84
Annex 1: Table of Evidence ............................................................................. 97
iii
Executive Summary
Aims and objectives:
1. This review of the empirical and theoretical literature was
commissioned by the Small Business Service (Department for Trade
and Industry). Our brief was to consider existing research that has
examined the causes and the impact of economic churn on productivity,
employment and ultimately economic growth. As well as considering
the effect of churn on markets, we were also asked to look at the
determinants of business start-ups and closures at the firm level; an
integral part of the churn process. Finally we were asked to identify
existing gaps in the literature and put forward methodologies for
addressing these.
Structure of report:
2. Following our introduction, our review begins with an overview of the
relevant theoretical arguments that offer insight into why churn takes
place. There are two reasons why churn is thought to raise productivity.
Firstly, through the competitive effect whereby entrants create an
environment in which incumbent businesses must „raise their game‟ in
order to remain in the industry. The second positive impact on
productivity is through the introduction of new technology. The most
notable body of literature is that which relates to the Schumpeterian
concept of „creative destruction‟, by which new firms enter markets
introducing new technologies and then displace older, less efficient
firms.
3. From the standard neoclassical theoretical position entry will take place
when firms identify a profitable opportunity and firms continue to enter
up to the point where the last entrant is breaking even. This standard
theory does not take into account asymmetries in information, sunk
costs, and other market imperfections. Instead, models of imperfect
competition are likely to be more realistic and mirror behaviour in the
„real world‟.
4. The importance of research in determining the level of creative
destruction is highlighted by macro economic models which emphasise
the importance of innovation. Macroeconomic conditions, such as the
interest rate, are also found to be important in determining the level of
churn. Industry models of entry and exit consider the relationship
between product diversity and entry both from an economic and
welfare perspective. Entry is likely to take place when firms feel they
have an alternative or improved product to bring to the market. A
number of these models highlight the importance of industry specific
characteristics in determining turnover rates. At the firm level, models
have been developed to analyse firm entry decisions and how they
survive by learning. Again, these models emphasise the importance of
industry specific factors such as market concentration.
iv
5. Our attention then turns more directly to the concept of economic churn
and its definition. Chapter 3 highlights a number of aspects that need to be taken into consideration when measuring economic churn. These differences are not only semantic but have real bearing on any comparisons that are made. We consider explicitly the importance of being clear at which level churn is considered. It could be argued that economic churn relates primarily to the entry and exit of businesses but a large proportion of the reallocation of resources takes place at the plant level within multiplant organisations. Another important factor to be aware of is that entry is not simply start-up, nor is exit simply business failure and we return to alternative modes of entry and to a lesser extent exit in Chapter 8. Finally, we highlight the importance of industry level and the time horizon in defining economic churn. When churn is considered at a relatively aggregate level, less reallocation between industries is observable. This needs to be considered when making comparisons between similar studies. The time period of analysis is similarly important, with short run improvements to productivity being less noticeable than long run effects. The theoretical chapter highlighted the importance of macroeconomic conditions and thus comparisons over time periods where there are differences in the economic cycle are likely to result in misleading conclusions.
6. Chapter 4 considers the measurement of entry and exit. From our
review of the literature we have found that there are a large number of ways in which churn can be measured. A „turn over‟ measure comprising of aggregated start-up and closure rates generally captures the overall churn effect. We find that in young industries, measures of net entry may also provide a clear indication of the competitive effect of entry. In more mature industries, where exit is likely to form a substantial proportion of the churn, a composite measure of turbulence might be more relevant. Ultimately, the most appropriate measure will depend on the purpose of the study and the industry conditions that are being examined.
7. Chapter 5 goes on to present empirical findings of churn rates from a
number of international studies. Our primary focus is on a recent study that goes to great lengths to harmonise definitions and time periods across countries. From this study it can be seen that there are higher levels of churn in service sectors compared to manufacturing sectors. The UK appears to have a higher recorded rate of churn in comparison to other countries. In other studies we see that there is also significant variation in UK regional churn rates that reflects the industrial profile of each region.
8. In Chapter 6 we consider the extent to which business churn leads to
improvements in the economy through employment, productivity and economic growth. Overall, creative destruction is viewed positively since it reallocates resources from less to more efficient uses in the economy. Given this, we would expect to see a positive impact on productivity in industries where churn rates are relatively high. Empirical research on this shows that the time horizon for analysis is
v
particularly pertinent for identifying the impacts from churn. Entry effects through start-ups are slow to generate higher levels of productivity but longer run panel studies have been more effective in finding positive productivity impacts. The relationship between churn and employment is also considered here. The entry of new plants is thought to be a major source of new employment opportunities and for this reason is often the focus of government policies to encourage start-ups, particularly in less successful regions. However, the literature emphasises that the short run productivity impact of entry is not likely to be substantial.
9. Through the churn process, firms and markets are thought to grow,
resulting in increased employment, productivity and welfare. It could be argued that creative destruction represents only a component of overall churn and that not all churn is beneficial. There are situations in which the level of churn may be considered excessive, for instance when entry crowds out incumbent firms leading to inefficient payment of sunk costs and no corresponding improvements to market performance.
Chapter 7 considers the costs associated with churn. The possibility that churn is not always positive and can result in the inefficient reallocation of resources might imply that an „optimal rate of churn‟ exists. This is an under-researched area in the literature and this review has been unable to identify an optimal rate of churn.
10. Chapter 8 returns to the issues of definition and measurement by
outlining the differences between the various modes of entry and exit. All forms of entry face common factors that influence the decision to enter but the relative importance of these is likely to favour one or other mode of entry. This chapter specifically considers diversification and changes in ownership as an alternative means of reallocating resources. Firms that choose to diversify are at an advantage over start-up entrants since they are better equipped to choose their preferred size of entry to test market conditions before committing substantial resources to the market. Diversifying firms may also have a brand name to exploit and are able to utilise resources and expertise from the parent company. As a result, entrants through diversification are generally larger than start-up entrants, and are more likely to enter concentrated markets. Diversifying firms are more likely to alter incumbent firm behaviour than new start-ups and have a greater competitive influence on incumbent behaviour.
11. Entry through changes in ownership such as mergers and takeovers
are both a form of market entry and exit. There are three basic reasons why acquisitions take place. Firstly, there are the potential synergies from subsuming an existing firm and streamlining duplicated departments to increase efficiency. Secondly, ownership changes may occur for reasons not associated with profit but in the pursuit of managerial objectives, such as the growth of the firm or managerial prestige. Finally, it could be that firms make mistakes in identifying potential synergy gains arising from a take over or merger and they do not actually materialise. From an exiting perspective, two competing theories exist. Firms may be taken over because they are performing
vi
badly, and thus takeovers are a form of managerial discipline. An alternative view is that well performing firms are targets for acquisition as the acquiring firm wishes to access greater operational efficiency by purchasing it. Entry through mergers and takeovers are also likely to be on a larger scale than start-ups and, as a means of churn, are less disruptive to workers and supply chains than business closures.
12. Chapter 9 is devoted to considering entry through new business start-
up and closures. These are the more prevalent modes of entry considered in this review, compared with diversification and changes in ownership. Our empirical focus is primarily on the UK, however we begin with some theoretical considerations about what constitutes a firm, drawing on the resource based model of the firm where the firm is viewed as a bundle of assets which are specific to that firm and not easily transferred through the market place. Our other key model is that of the evolutionary approach, which views the firm more as a means of facilitating technological change and market knowledge. Policy recommendations for the support of start-ups are likely to differ, depending on the model that dominates in the industry.
13. We consider business start-up rates and find that these vary
considerably across the UK, and by region. Existing research focuses almost exclusively on the manufacturing sector and thus there are problems with drawing economy-wide conclusions on the basis of this evidence alone. A number of key factors emerge from our
consideration of the literature. We find that it is widely recognised that entry rates are highly correlated with exit rates, which is partly due to survival rates of new entrants being relatively low. We also see that industry churn rates are broadly similar across countries, suggesting the importance of industry specific factors, in additional to institutional factors. This suggests that industry specific sunk costs act as a deterrent to entry, and consequently also act as a barrier to exit as well.
14. Given the high correlation between entry and exit, we go on to consider
the role of entrepreneurship in start-ups, looking at the empirical evidence to establish what factors contribute to a successful entrepreneur. We find that both environmental factors (such as the region – population levels, local government policy) and socio-
economic factors (such as wealth, family background and education) have significant influences on whether individual choose to start-up their own business. The main motivation for starting ones own business is reported as being „to be ones own boss‟. Implementing a new idea i.e. innovation, was a much less popular response. This suggests that a large number of entrants are not bringing an innovation to the market and may account for the high number of exits due to displacement of one another.
15. The literature looks at industry conditions as a key determinant of start-
up. We find that the level of industry concentration and various barriers to entry are all thought to contribute significantly to determining the number of start-ups in an industry. The greater the level of observed profitability in an industry the more likely entry is to occur, subject to
vii
barriers to entry prevalent in the market. This suggests that the less concentrated the industry is, the higher the probability of entry occurring. Industry life cycle effects are also likely to influence the proportion of start-up entry. The importance of region is again emphasised and is particularly important with respect to the policy implications of churn. These are considered in greater detail in Section 9.5.
16. In Section 9.3 we go onto to consider the literature that looks at the
importance of innovation in business start-ups. The link between start-ups and innovation is discussed extensively in the small business literature, and we find that there is support for the „u‟ shaped relationship between firm size and innovation, whereby firms that are small and large are the most likely to innovate.
17. A recurring theme throughout the literature on churn is the high
mortality rate of new entrants. In Section 9.4 we examine the relationship between survival and exit, looking particularly at the factors which are likely to assist plant survival. We see that size and age are found to be especially important in survival and consider the validity of Gibrat‟s Law that asserts that growth is independent of size. We find that a large body of empirical literature has found that the likelihood of firm survival is not independent of firm size.
18. Chapter 9 considers the UK current enterprise policy. It has recently
been argued that there is a lack of transparency in the objectives of small business policy. Our findings in relation to the costs of churn suggest that encouraging excessive entry is not likely to result in higher employment and growth, but is likely to raise the exit rate correspondingly. Assistance to small business perhaps needs to focus on a nurturing role, whereby government aims to create a conducive environment for all enterprises. However a potential problem with the increased use of relatively „soft‟ intervention is that its impact is much more difficult to assess.
19. Chapter 10 considers approaches for analysing the aggregate impact
of churn. This chapter highlights a major gap in existing literature in that the vast majority of studies use a bottom-up approach with the use of micro data. One study that considers a top-down approach looks at the extent to which the gap between actual and optimal industry structure impacts on growth.
20. The report concludes with a section that puts forward a number of
suggestions for further work in this area. Most notably, the extension of the micro data based analysis of the service sector is essential to gain an insight into the impacts of entry and exit within this sector. In addition, this area of churn and productivity could be considered more thoroughly using a top-down approach with industry level datasets.
viii
1. Introduction
The continual entry and exit of plants and firms is crucial to the growth of an economy. This turbulent process is known as economic churn and
is thought to impact on industry productivity through increased competition leading to greater cost efficiency. It can also facilitate creative destruction whereby more innovative firms enter the market, replacing existing incumbent firms using current technologies.
The role of dynamic markets in reallocating resources through the entry and exit of firms has been explored extensively in the industrial organisation literature, and various aspects of entry and exit have been considered in the small business and the regional literature. This review draws on these areas in an attempt to explore the channels through which entry and exit takes place and the influence this dynamic process has on productivity at the industry and national level. As well as the positive effects associated with the replacement of exiting firms, it is recognised that there are costs associated with the reallocation of resources such as the destabilising effects caused by redundancy. There is a trade-off between the benefits churn brings and the costs associated with the reallocation process and this will be explored further within the review.
The purpose in this literature review is to consider economic churn in detail. The review begins in chapter 2 with a discussion of the relevant theoretical models which have implications for the nature of entry and exit. On close inspection of the literature, we find that definitions are central to our understanding of what is meant by the phrases „churn‟, „entry and exit‟ and „creative destruction‟. Particularly, a clear understanding of the definition of churn or entry and exit is essential for meaningful cross-country comparisons to be made. In chapter 3 we provide a detailed explanation of the terminology used in the literature.
Having considered the definitional issues, chapter 4 then considers measurement issues and the different approaches for calculating churn. In chapter 5, the review looks at the existing empirical evidence of churn from a large range of countries. In chapter 6 we are specifically concerned with presenting a complete picture of the role that economic churn plays in the economy by considering its impact on productivity, employment and ultimately GDP growth. Chapter 7 identifies that not all churn raises productivity and employment and, where churn is „noisy‟, it may be an inefficient use of resources. We therefore consider the costs that arise from business churn. Our findings from the literature suggest that it has not been possible to robustly calculate an optimal level of churn, and that churn rates are largely determined by industry and market specific conditions.
Chapter 8 considers diversification and changes in ownership as alternative means of transferring resources and as alternative modes of churn to start-up and closure. These may have a larger impact on the industry than just focusing on start-ups. In chapter 9 we consider the role of start-ups and closures in the UK and identify the obstacles that they face and the policy environment in which they operate. The importance of start-ups in the process of churn is not so evident in the short run since start-ups face high failure rates and represent a small proportion of total market share. However, the impact of start-ups should not be underestimated as they bring innovations to the market and in the longer run, expand their market share. Therefore, the time dimension is a very important factor to consider when assessing the impact of churn.
In chapter 10, we highlight the scarcity of studies that take a top-down approach to the impact of churn instead of the current focus on micro data. The use of top down approach enables the use of richer sources of data than are available at the micro level and ensures a greater consistency across space and time. In the final chapter of the review we briefly consider the main gaps in the existing evidence and highlight areas for further research. In particular, we identify the poor coverage analysis relating to the service
2