The Number and Size of Firms: Why So Big a Difference? Štefan Bojnec Ana Xavier The number of firms and their size are analysed for the Slovenian manufacturing sector on the basis of the firm-level evidence of the Business Register of Slovenia virtually representing all the firms in activity. A remarkable difference is found between the number of the registered manufacturing firms and the number of firms with employment. The increase in the number of all registered firms is remarkable, but it is less so for the number of firms with employment, suggesting that many more firms were being registered than were in reality economically active. The large majority of newly registered firms during the 1990s were firms without any recorded employment. Whilst the number of firms increased, the number of employees declined, the average manufacturing firm size measured by employees per firm declined. Private firms constitute the vast majority of the firms in activity at end of the 1990s and afterwards. Introduction The rapid increase in the number of firms serves as an indication that markets are becoming more competitive. As in several other Central and Eastern European (cee) countries, the number of firms in Slovenia increased substantially throughout the 1990s early 2000s. The considerable increase in the number of firms is a result of policy changes associated with relatively low capital and registration requirements, easing the opportunities for registering and entry of new registered firms, as well as Dr Štefan Bojnec is Associate Professor of Economics at Turistica -College of Tourism Portorož, University of Primorska, Slovenia. Dr Ana Xavier is an Official Administrator, d g Employment and Social Affairs, European Commission, and an Associated Researcher at licos - Centre for Transition Economies, Katholieke Universiteit Leuven, Belgium. This research was undertaken with support from the European Union’s Phare- ace Programme 1998 and from the Slovenian National Science Foundation. The content of the article is the sole responsibility of the authors and does not necessarily represent the views of the Commission, its services or any other institution. Managing Global Transitions 2 (2): 89–105 90 Štefan Bojnec and Ana Xavier restructuring, bankruptcy and organisational transformation within the large manufacturing enterprises, and similar processes that took place during transition to a market economy. Overall economic deregulation has allowed new firms to enter the markets, some of which were successful and have been in activity for a long period of time and some of which failed and exited the market (e. g. ebrd 2001). Several new private de novo firms have been established, and also many large manufacturing enterprises were split or reorganised into several parts. The increase in competitive pressure has been induced by the entry of new firms and by trade liberalisation. The bankruptcy procedure has played an important role in imposing hard budget constraints and straightening financial discipline for exit of loss making enterprises. While some firms exited, particularly in the most recent years, as a result of increased competitive pressures in firm output markets and institutional and policy changes in line with those of the European Union (eu), firm entry still offset firm exit (Bojnec and Xavier 2004). Hence, an increase in the number of firms can be acknowledged albeit with different intensities across different branches. In this paper we look more in-depth into the growth of the number of firms aiming to make a distinction between the growth in the number of job creating firms and that of the firms without any recorded employment. Indeed, one of the most striking features of Slovenia’s transition is the remarkable increase in the number of firms recorded by the official statistics in the Business Register of Slovenia (brs). However, there is a considerable difference in the number of registered firms and the number of economically active firms. This ‘vacuum’ in the number of firms due to several ‘empty’ firms has so far been largely neglected. We draw attention to this fact by dividing firms into active firms (‘non-zero employment firms’) and inactive firms (‘zero employment firms’) according to their employment. While the latter are by far the most important in the brs, any serious econometric analysis of the firm dynamics can only be conducted on the basis of the economically active firms with at least reasonably good evidence for employment, financial and some other performance indicators (Bojnec and Xavier 2004). Therefore, this paper aims to analyse the extent of the gap between the two categories of firms and its impact on the firm numbers and size, as well as providing an answer to the question of why there is so great a difference between the larger number of registered and the smaller number of economically active firms, which causes differences in the firm size. We Managing Global Transitions The Number and Size of Firms: Why So Big a Difference? 91 look at the extent and the evolution of manufacturing firm demography on the basis of the firm-level information of the brs obtained from the Statistical Office of the Republic of Slovenia (sors) virtually representing all the firms in activity at any point between 1987 and 2000. Most manufacturing branches in Slovenia and - throughout the 1990s - experienced rates of firm entry that are greater than firm exit (Bojnec and Xavier 2004). Most of them also experienced labour shedding associated with an initial increase in unemployment and, in the mid-1990s, with an outflow of labour into regular and early retirement. As a result, the average size of manufacturing enterprises in terms of employment per firm declined during that period. There are, however, differences in manufacturing enterprises according to firm ownership. Private firms are responsible for the greatest difference in the number of firms. Some of them are part-time self-employed firms, but several of them are ‘empty’ firms without any employment. The Number of Firms To analyse the dynamics in the number of firms and their size in the manufacturing d sector, we use firm-level information provided by the brs obtained from the sors. This data set provides us information on firm identification (id), nace sector, employment and firm ownership. The distinctions between firms are made on the basis of the firm id used as the criteria to identify whether a firm has stopped its economic activity (if id is no longer in the sample), has started activity (id is not previously in the sample) or is still in activity (if 1 d is still in the sample).1 Note that there were some cases when a firm appeared to have exited in one year but in fact reappears later on as a ‘survivor’, with the same id number, which indicates that when it was first considered an ‘exitor’ it was potentially due to misreported information. These cases were re-coded as ‘survivors’. We distinguish between zero- and non-zero employment firms. The non-zero employment firm is a firm whose employment is a certain positive number equal to or greater than 1. The zero-employment firm is one whose employment in the dataset is recorded as zero. The latter situation might represent part-time self-employed persons (e. g. also employed somewhere else or retired) or firms whose operation has not started or already stopped, but the firm is still in the brs. Very often they correspond to firms who have formally registered but whose activity has not started in reality. As a result of institutional changes, which made firm entry easier with relatively low initial capital and other registration Volume 2 · Number 2 · Fall 2004 92 Štefan Bojnec and Ana Xavier requirements in the early 1990s, and the economic transformation leading to firm restructuring, spin-offs and by-pass firms, the number of the manufacturing firms in Slovenia increased from 1,614 in 1987 to 8,783 in 2000 or by 444% (Figure 1). Interestingly, the rapid increase in the number of manufacturing firms in Slovenia is due in particular to the substantial increase in the registered number of firms which do not record any employment or the number of ‘zero’ employment firms. The number of manufacturing firms with ‘zero’ employment increased from 241 in 1987 to 6,992 in 1998 or by 2,801%, slightly declining thereafter. There was a window of opportunity, which opened and allowed the establishment and setting up of new private or de novo firms. There is also a positive development pattern for the firms with recorded employment, but their increase is more modest. They increased from 1,373 in 1987 to 2,132 in 2000 or by 55%, which is still a considerable increase in their numbers. When looking at firm ownership we can only consider the firms for which ownership information was clearly stated (Tables 1 to 3). The focus of our analysis is on the number of manufacturing firms by their ownership. The total number of firms is divided by ownership category looking at the whole sample and the two sub-samples controlling for firm employment. The ‘zero’ employment firms are those which do not record any employment. The ‘non-zero’ employment firms are those which record a positive (>o) number of employees. We use the classification of the firms provided by the sors, which in the brs classifies the firms according to the prevailing firm ownership in the following four categories: socially owned enterprises, mixed enterprises, cooperatives or enterprises owned in the majority by cooperatives, and privately owned enterprises. Since in several cases the firm ownership in the brs is not clearly identified, we introduce an additional category of the ‘not-identified’ firm according to its ownership. The relatively high number of firms which are considered ‘not-identified’ according to firm ownership occurred particularly during the first years of the 1990s. The number of socially owned manufacturing firms initially declined, increased between 1992 and 1994, and declined again afterwards. This development is a mixture in development of socially owned zero and nonzero employment firms. The drop in the number of socially owned enterprises in the years 1990,1991, and 1992 is reflected in an increase in the number of ‘not-identified’ enterprises. Since then, the number of ‘not-identified’ enterprises has declined. This has been accompanied by an increase in the number of mixed zero and non-zero employment firms. Managing Global Transitions The Number and Size of Firms: Why So Big a Difference? 93 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 o y = 637 09x + 654.79 r˛= °.9!49 -">< XT' y / / / y = 560.60x -164.26 R˛= 0.8747 y'7 y = 80.71x + 985.46 R˛= 0.751 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Total number of firms, ‘Zero’ employment firms, ‘Non-zero’ employment firms, linear trend linear trend linear trend Figure 1: Number of registered firms in Slovenian manufacturing, 1987–2000 Source: Authors’ analyses on the basis of the data from the brs. These developments suggest that at the beginning of the 1990s there was a period when an important reorganisation of socially owned enterprises took place in agreement with the changes in the regulatory and institutional environment. The number of mixed firms has approximately tripled since 1990. It increased steadily until 1998, but slightly declined afterwards. The growth of the non-zero employment mixed-owned firms was fastest than that of the zero employment mixed-owned firms. The relatively small number of cooperatives increased with a high prevalence of zero employment cooperatives over time. The most considerable is, however, the increase in the number of private firms. The number of Volume 2 · Number 2 · Fall 2004 94 Štefan Bojnec and Ana Xavier Table i: Number of total registered firms by ownership in Slovenian manufacturing, 1987–2000 Year Social Not identified Mixed Co-operative Private 1987 I614 0 0 0 0 1988 I46I 0 0 0 0 1989 1418 0 0 0 0 1990 661 191 106 4 1638 1991 914 79 222 6 2436 1992 1012 37 268 8 3731 1993 1041 12 308 15 4882 1994 1036 12 308 18 5234 1995 962 22 301 19 5448 1996 800 26 324 26 5226 1997 719 9 384 58 7288 1998 598 9 400 62 7692 1999 529 16 387 62 7778 2000 460 24 349 62 7033 Source: Authors’ analyses on the basis of the data from the brs. private manufacturing firms with recorded employment increased from 5 in 1990 to 1,747 in 2000 (Table 2). However, the most striking is the fastest growth of private firms with zero employment from 1,633 in !99° to 6,320 in 1998 or by 287%, but with a slight reduction in their number afterwards. Tables 1 and 3 clearly illustrate that a large number of private manufacturing firms were set up between 1990 and 1993, but many of them did not create employment. Several private firms remained inactive for job creation even for a longer period. In 2000, 91.2% of ‘zero’ employment firms were private manufacturing firms. Hence the main difference in the number of the registered firms and the number of firms with employment is due to a large number of private firms with ‘zero’ or no recorded full-time employment. More specifically, Figure 2 compares our results with the recorded evidence in the Statistical Yearbook of Slovenia (syslo). According to our results, since the Law on Enterprises entered into force in 1989, more than 70% of all manufacturing firms were firms without any recorded full-time employment. In 1997, this share was over 80%, but with an important decline thereafter. The syslo, however, did not record any dis- Managing Global Transitions The Number and Size of Firms: Why So Big a Difference? 95 Table 2: Number of firms with ‘non-zero’ employment by ownership in Slovenian manufacturing, 1987–2000 Year Social Not identified Mixed Co-operative Private 1987 1373 0 0 0 0 1988 1252 0 0 0 0 1989 1106 0 0 0 0 1990 611 139 53 4 5 1991 788 38 162 6 8 1992 854 14 194 6 259 1993 831 2 222 9 352 1994 769 1 215 10 492 1995 626 10 201 8 582 1996 433 20 222 10 834 1997 304 4 262 14 1214 1998 213 5 270 17 1372 1999 154 8 256 17 1675 2000 116 11 243 15 1747 Source: Authors’ analyses on the basis of the data from the brs. crepancies in the number of firms until the mid-1990s. In the years 1997– 1998, the syslo recorded a discrepancy between the number of manufacturing firms and the number of firms with full-time employed persons who are insured at the pension and disability insurance and health insurance of around 43–44%, which is less than our calculations indicate. Finally, we provide a comparison between the number of active firms with recorded payments through resident accounts and final accounts held by the Agency of the Republic of Slovenia for Payments, and the number of registered firms in the brs. According to this comparison, around 70% of registered firms are financially inactive firms. This finding is much closer to our result when comparing the number of firms and the number of zero employment firms. These comparisons suggest that the increase in the number of economically active manufacturing firms in terms of employment or in terms of financial transactions was not as substantial as initially, particularly until the mid-1990s, recorded by the official statistics on the basis of the brs based on different (inconsistent) registers. However, the question remains of how big this gap is between the registered and economically active manufacturing firms and what can explain Volume 2 · Number 2 · Fall 2004 96 Štefan Bojnec and Ana Xavier Table 3: Number of firms with ‘zero’ employment by ownership in Slovenian manufacturing, 1987–2000 Year Social Not identified Mixed Co-operative Private 1987 241 0 0 0 0 1988 209 0 0 0 0 1989 312 0 0 0 0 1990 50 52 53 0 1633 1991 126 41 60 0 2428 1992 158 23 74 2 3472 1993 210 10 86 6 4530 1994 267 11 93 8 4742 1995 336 12 100 11 4866 1996 367 6 102 16 4392 1997 415 5 122 44 6074 1998 385 4 130 45 6320 1999 375 8 131 45 6103 2000 344 13 106 47 5286 Source: Authors’ analyses on the basis of the data from the brs. it. Some ‘zero’ employment firms might be self-employed firms not obligated to conduct payments through recorded resident accounts to the Agency for Payments. However, what the data appear to suggest is that the number of part-time self-employed firms was of much less importance than the number of firms without any recorded employment.2 This clearly confirms that, to a great extent, ‘zero’ employment firms were ‘empty’, economically inactive firms, whose number has started to decline since 1998 and more recently with the ‘cleaning’ of the brs. Average Firm Size The average size of manufacturing firms is analysed on the basis of the number of employees per firm (mean employment per firm). Labour shedding and retirements during economic transition and firm transformation in the Slovenian manufacturing sector were significant, and consequently the total number of employees in the manufacturing sector declined from 321,945 in 1987 to 177,121 in 2000, or by 45%. The greatest decline in employment in Slovenia occurred between 1987 and 1993. The most considerable decrease in employment in the manufacturing sector occurred in 1990 when the rate of employment growth amounted to (mi- Managing Global Transitions The Number and Size of Firms: Why So Big a Difference? 97 80% 60% ‘Zero’ SORS 40% 20% 0% AP 1987 1988 1987 1988 1987 1988 Figure 2: Discrepancies for the number of inactive manufacturing firms, 1997–1998 Notes: ‘Zero’ - percentage of‘zero’ employment firms in total number of registered manufacturing firms; sors - percentage of inactive firms with no employed person in total number of manufacturing firms as recorded by sors; ap - percentage of active firms with recorded payments through resident accounts or final accounts by the Agency of Payments in the number of registered firms in the brs. Source: Authors’ analyses on the basis of the data from the brs, syslo 1997, and syslo 1998. nus) –17.2%. Since 1997 the rate of employment growth has been positive, indicating a slight increase in the total number of employees in Slovenian manufacturing. As the number of employees declined and the number of firms increased at the same time throughout this period, it is logically the case that the average number of employees per firm declined. This is clearly revealed by the empirical results in Figure 3. Between 1987 and 2000 the mean number of employees per registered firm in the manufacturing sector declined from 199.5 employees to 22.3 employees. While at the end of the 1980s, an increase in the average size of the firm in terms of the employee number per firm is recorded, later during the 1990s the average size of the firm declined steadily. The most considerable difference in the average size of the firm occurred in 1990 due to the rapid increase in the number of ‘zero’ employment firms. This clearly indicates that the average size of the firm in terms of employment per firm is biased considerably towards the firms without employment. Some of them are new entries in a form of self-employment, while a large number of them are ‘empty’ firms, which exist only on a paper as a result of the transformation of existing enterprises and institutional changes, which made firm Volume 2 · Number 2 · Fall 2004 98 Štefan Bojnec and Ana Xavier 300 200 100 -^-« y = 0.2854x ˛ - 20.925x + 302.77 R˛ = 0.8367 y = 2.283x ˛ - 49.456x + 278.7 R˛ = 0.9034 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 -------- Employment per registered firm, polynomial trend --------Employment per ‘non-zero’ firm, polynomial trend Figure 3: Mean employment in manufacturing, 1987–2000 Note: The values regard numbers of workers per firm. Source: Authors’ analyses on the basis of the data from the brs. registration easier. Some large enterprises were also left ‘empty’ by establishing a by-pass firm. To exclude this potential bias towards the ‘empty-zero’ employment firms, and to estimate the sensitivity of the results, we present also the results of the average size of manufacturing firm in terms of employment per ‘non-zero’ employment firm. It is clearly visible from Figure 3 that the average size of the ‘non-zero’ employment firm is much greater than the average size in the whole sample of firms. The former declined from 234.5 employees per firm in 1987 to 83.1 employees per firm in 2000, while the latter declined from 199.5 to 22.3 during the same period, clearly showing that the gap, caused by the bias in the development of ‘zero’ employment firms, does exist, but after having first widened it has been narrowing. The average size of the ‘non-zero’ employment manufacturing firm increased in 1988,1989 and 1990 when most of large socially owned enterprises were not being transformed and restructured. Some of them continued to grow. However, the average size of the ‘nonzero’ employment firm declined considerably in 1991 and in 1992. This is consistent with some fundamental institutional and policy changes leading to an extensive process of firm organisational transformation and firm restructuring imposed by the institutional and policy changes. Most large manufacturing enterprises underwent a process of reorganisation and restructuring in preparation for the process of privatisation. Later, the average size of the ‘non-zero’ employment firm continued to decline steadily. A convergence process between the two-trend lines (for zero and o Managing Global Transitions The Number and Size of Firms: Why So Big a Difference? 99 non-zero employment categories) is clearly illustrated for the most recent years suggesting that some cleaning process is going on among the registered ‘zero’ employment firms. Factors Affecting Changes in the Number of Firms and Their Size As said, the main reason for the large difference in numbers of zero and non-zero employment firms, and consequently in the average firm size, are ‘empty’ firms, that is, the vast majority of zero employment firms are economically inactive firms without employment or without sales that exist only on paper and in the brs. Many are private firms. Between 1989 and 1993, and to a lesser extent later, several firms were set up, which have never been economically active either in terms of generating jobs or in sales. This gap consisting of zero employment firms overestimates the growth of newly established firms. A certain cleaning of the brs can be noticed after the year 1998, when the peak in the number of private ‘zero’ firms was achieved. During the period analysed here, several changes have taken place within the institutional and legal frameworks and in governmental policies, which have affected the dynamics of firms. Among the main institutional and policy changes related to the dynamics of manufacturing firms are: requirements for setting up firms, bankruptcy and bankruptcy procedures, government rehabilitation policies associated with potential state subsidies, competition and trade policies. LIQUIDITY, LOSSES, BANKRUPTCY AND BANKRUPTCY PROCEDURES A body of literature has developed on the role of initial conditions on transformation, and later recovery and economic growth (e. g. Aghion et al. 1994; De Mello et al., 2001; Falcetti et al., 2002). One often recorded constraint for Slovenian firms is the delay in receiving payments and the associated liquidity problems.3 The number of firms and the share of employment in illiquid companies with blocked giro accounts for more than five days was quite high in the first half of the 1990s (imad 1994,21). As in other cee countries, a large fraction of enterprises was faced with financial difficulties. While in the initial stage of transition most illiquid enterprises were large enterprises, later on, among the illiquid pool of enterprises, there also appeared many medium size and small firms with a smaller number of employees per illiquid firm. Volume 2 · Number 2 · Fall 2004 loo Štefan Bojnec and Ana Xavier The legal and regulatory basis for bankruptcy and bankruptcy procedures is one of the most crucial elements for efficient market selection process in market economies (e. g. Gray 1993; Coricelli and Djankov 2001; Maskin and Xu 2001). The legal and regulatory environment for bankruptcy and bankruptcy procedures in Slovenia has passed through different stages, which were largely related to the various bankruptcy laws and their implementation. Companies’ bad liquidity position and financial disorder were an important systemic problem (e. g. Zizmond !993). The bankruptcy law represents the key regulatory environment for the bankruptcy of enterprises as a process in which the interests of owners, borrowers and employees should be matched. Protection of employment or labour hoarding has been argued from public or national interest as a means to overcome the financial crisis and to sustain employment. Throughout the 1990s some manufacturing enterprises did go bankrupt, and falling employment appears to have been more the result of enterprise bankruptcy than of enterprise restructuring (ebrd 1998). Some of these enterprises were later re-established, but as a rule - particularly in labour intensive enterprises - they reduced the level of employment. Among capital-intensive enterprises, they were often heavy industry enterprises and loss-making enterprises during the longer period in the 1980s.4 While for labour intensive enterprises it was common to encounter labour shedding, for capital-intensive enterprises it was common to have the assets written off or reduced. Some training/re-education programmes to re-qualify and train workers were developed and directed at labour intensive branches. However, due to the relatively high wages in Slovenia - a fact that alters international competitiveness - several labour intensive activities are still under the pressure of having a relatively low value-added to pay high wages and to compete in the international markets. Among such branches are the textile and footwear industries. ENTERPRISE TRANSFORMATION AND GOVERNMENT RESTRUCTURING PROGRAMMES The State Development Fund was an enterprise-restructuring agency for the restructuring of large-loss-making enterprises. The Fund took over several enterprises in financial difficulties for possible restructuring. In 1997 it was transformed into the Development Corporation of Slovenia (dcs) (Slovenska razvojna družba - srd) for the financing and restructuring of enterprises that had not yet been privatised, as well as priva- Managing Global Transitions The Number and Size of Firms: Why So Big a Difference? 101 tised enterprises in financial distress (ebrd 1998). Large-scale enterprises in the aluminium, steel, and oil sectors were included in the rehabilitation process, relaxing firms’ budget constraints, and in providing them with subsidies. The prevalence of soft-budget constraints was intended to mantain employment levels and lead to a gradual restructure of the enterprise. In some enterprises development centres were established or re-established (e. g. wood and textile industry) aiming at employee retraining and enterprise adjustment towards a greater ability to compete on developed western markets. With the abolishment of the dcs, the enterprises in the dcs are being transferred to privatisation investment companies, pension funds and some other agencies. COMPETITION POLICY Competition policy may allow firms easier entry into the market and exit from it. It is recognised by the existing literature that healthy competition increases competitive ability and leads to greater competitiveness (e.g. Bresnahan and Reiss, 1991; Aghion at al., 1997). Unfair competition via entrepreneurial restriction of competition reduces competitive ability and hence is a deviation or a violation of good business or trade practices. This can be due to market power by individual participants, systemic distortions, and distorted economic policies. Competition policy and the protection of competition aim at allowing full existence of market competition, ensuring market discipline, and preventing of unfair competition in the interests of society, companies and consumers. Competition policy covers both the areas of anti-trust regulation (prohibition of monopolistic agreements and abuse of the monopolistic position) and regulation of (unfair) competition (prohibition of unfair competition), which is harmonised with eu regulations and their implementation. TRADE POLICIES During the 1990s, four events were of major importance to Slovenian trade, which have had implications on the restructuring, entry and growth of firms. First, there was the breakaway from the former traditional markets in the former Yugoslav republics. This caused demand shock for some enterprises previously largely selling products to these markets. Second, a new free trade initiative developed in the region, which resulted in the Central European Free Trade Agreement (cefta). This was beneficial for sales and growth of several manufacturing firms. Volume 2 · Number 2 · Fall 2004 102 Štefan Bojnec and Ana Xavier Third, at the end of 1994, Slovenia became a member of the General Agreement on Tariffs and Trade (gatt) and one of the founding members of the World Trade Organisation (wto). With the wto membership, trade measures have been transformed into more transparent and less discretionary trade policies. This has had more indirect effects on firms. Fourth, the widening and deepening of the East-West European integration was stipulated by the Association Agreements with the eu, and deepened through the negotiation process and the eu membership on the ist of May 2004 with the adoption of the entire acquis commu-nautaire. Manufacturing products constitute the most important item in Slovenian trade. While Slovenian firms had already developed some ways of cooperation with eu firms during the old system, these initiatives and adjustments to the eu membership have further boosted and created the growth of Slovenian merchandise trade with the EU-15. OTHER GOVERNMENT POLICIES AND MEASURES Among other government policies are fiscal and budgetary policies. More specifically, taxation policies can provide incentives or disincentives for the setting up and growth of newly established firms. Among important measures are also policies regarding the banking system, which can provide incentives or disincentives for firm development, serving them by providing loans to firms under internationally comparable competitive loan conditions. Conclusion This paper analyses the evolution of the total number of Slovenian firms establishing a clear distinction between firms with employment (‘nonzero’ employment firms) and firms without any recorded full-time employment (‘zero’ employment firms). In doing so the paper observes a large gap between the large total number of registered firms and the much smaller number of firms with employment. Thus, the crucial finding of this paper is that the recorded extremely fast growth of the total number of manufacturing firms in Slovenia throughout the 1990s was exaggerated. It is clearly illustrated that the number of manufacturing firms increased, but less substantially in real economic sense than was initially recorded by statistics. The increase in the number of economically active firms (i. e. ‘non-zero’ employment firms) was less substantial (albeit considerable) than the total number of firms thus indicating that a vast number of registered firms were never active in economic sense Managing Global Transitions The Number and Size of Firms: Why So Big a Difference? 103 throughout the 1990s. Indeed, particularly great was the increase in the number of ‘zero’ employment firms which to a large extent are ‘empty’ economically inactive firms that exist only on paper in the brs, with the exception of some part-time self-employment firms. These ‘empty’ firms are of statistical nature, closely associated with the institutional deregulation associated with the transition process, allowing for easier firm’s registration at relatively very low initial capital and other registration requirements. It is necessary to mention that the increase in the number of ‘nonzero’ employment firms is related to a considerable increase in the number of privately owned firms. Indeed, several large manufacturing enterprises were transformed into various organisational units, and different newly established private firms were set up, which are typically smaller. The increase in the number of manufacturing firms and the reduction in the number of employees led to the decline in the firm size (i. e. in the number of the employees per firm) during the 1990s, which is however smaller if the ‘zero’ employment firms are disregarded. Note that a recent convergence process is taking place between the total number of registered firms and the total number of ‘non-zero’ employment firms, leading to the convergence of firm size measured by employment per all registered firms and by employment per ‘non-zero’ employment firms. This suggests that some ‘cleansing’ process is ongoing among the firms without recorded full-time employment or ‘zero’ employment firms. With the Slovenian harmonisation of the institutional environment and implementation of policies in line with the eu, the market selection process in terms of firm entry and exit is again more dynamic. Since 1997, the rate of employment growth has been positive indicating a slight increase in the number of employees in Slovenian manufacturing. Further deregulation, strict implementation of the rules of competition and hence harder budget constraints may be expected to render the market selection process similar to what can be observed in the developed market economies. Notes 1. Note that changes in the firm’s name are not considered, but only changes in the firm identification number (id). The comparison between the name of the firms and the id of the firms suggests that many Slovenian firms (about one third) changed their names, but continued operating within the same firm’s id. Volume 2 · Number 2 · Fall 2004 104 Štefan Bojnec and Ana Xavier 2. For the difference in the number of firms, the explanation in the syslo (1987–2003) seems to be a bit weak. It is argued that the discrepancy in data between the brs and the Statistical Register of Labour Force (srlf) arises due to the fact that the brs comprises also inactive physical persons, and the srlf comprises physical persons who have not yet been registered in the brs, but have pension and disability insurance and health insurance. It is not clear by whom they are included in pension, disability, and health insurance. It is also not mentioned, that probably the main reasons for the difference are several newly registered firms by different kinds of employees, retired, students and similar persons, although several of these registered firms have never been active or finally stopped their activities, yet the firm’s registration remains in the brs. The latter argument is also more in line with the evidence from the register of resident accounts and final accounts at the Agency of Payments during the 1990s. 3. Inter-enterprise arrears and delay payments have been common during transition. For example, according to firm survey results recorded by Bojnec (2002), the typical delay payment period was about 43 days. 4. For similar problems in other cee countries see Hughes and Hare (1992), Aghion at al. (1994), Carlin at al. (1995), Earle (1997), Caves (1998), and Coricelli and Djankov (2001). References Aghion, P., O. Blanchard, and R. Burges. 1994. The behaviour of state firms in Eastern Europe, pre-privatisation. European Economic Review 38:1327–49. Aghion, P., M. Dewatripont, and P. Rey. 1997. Competition, financial discipline and growth. European Economic Review 41:797–805. Bojnec, Š. 2002. Payments, insolvency and finance during economic transformation: Slovenia on the way to European Union accession. Europe-Asia Studies 54 (2): 277–97. Bojnec, Š., and A. Xavier. 2004. Entry and exit in transition economies: The Slovenian manufacturing sector. Post-Communist Economies 16 (2): 191–214. Bresnahan, T. F., and P. C. Reiss. 1991. Entry and competition in concentrated markets. Journal of Political Economy 99 (5): 977–1009. Caves, R. E. 1998. Industrial organization and new findings on the turnover and mobility of firms. Journal of Economic Literature 36:1947–82. Carlin, W., J. V. Reenen, and T. Wolfe. 1995. Enterprise restructuring in early transition: The case study evidence from Central and Eastern Europe. Economics of Transition 3 (4): 427–58. Coricelli, F., and S. Djankov. 2001. Hardened budgets and enterprise re- Managing Global Transitions The Number and Size of Firms: Why So Big a Difference? 105 structuring: Theory and an application to Romania. Journal of Comparative Economics 29 (4): 749–763. De Melo, M., C. Denizer, A. Gelb, and T. Stoyan. 2001. Circumstances and choice: The role of initial conditions and policies in transition economies. World Bank Economic Review 15 (1): 1–31. Earle, J. S. 1997. Industrial decline and labour reallocation in Romania. William Davidson Institute Working Paper 118. EBRD. 1998. Transition Report 1998. London: European Bank for Reconstruction and Development. EBRD. 2001. Transition Report 2001. London: European Bank for Reconstruction and Development. Falcetti, E., M. Raiser, and P. Sanfey. 2002. Defying the odds: Initial conditions, reforms, and growth in the first decade of transition. Journal of Comparative Economics 30 (2): 229–50. Gray, C. W 1993. Evolving legal frameworks for private sector development in Central and Eastern Europe. World Bank Discussion Papers 209. Hughes, G. A., and P. G. Hare. 1992. Industrial policy and restructuring in Eastern Europe. Oxford Review of Economic Policy 8 (2): 82–104. iMAD. 1994. Slovenia: Economic trends in 1994 and the outlook for 1995– 1998; Spring report. Ljubljana: Institute of Macroeconomic Analysis and Development. Maskin, E., and C. Xu. 2001. Soft budget constraint theories. Economics of Transition 9 (1): 1–27. syslo. 1987–2003. Statistical Yearbook of Slovenia. Ljubljana: Statistical Office of the Republic of Slovenia. Zizmond, E. 1993. Slovenia: One year of independence. Europe-Asia Studies 45 (5): 887–905. Volume 2 · Number 2 · Fall 2004