• Keine Ergebnisse gefunden

Labour Market Policy and Employment

N/A
N/A
Protected

Academic year: 2022

Aktie "Labour Market Policy and Employment"

Copied!
174
0
0

Wird geladen.... (Jetzt Volltext ansehen)

Volltext

(1)

discussion paper

FS I 93-201

Self Employment and Labour Market Policy in the European Community*

Nigel Meager* •«

* Report prepared for the Commission of the European Communities (DGV). Final Report (draft)

Prepared with the collaboration of C. Degen, A. Deberdt, H. Hocker, und L. Linke

** Wissenschaftszentrum Berlin fiir Sozialforschung (WZB)

and

Institute of Manpower Studies (IMS), University of Sussex, Brighton, Great Britain

Marz 1993

ISSN Nr. 1011-9523

Research Area:

Labour Market

and Employment

Research Unit:

Labour Market Policy and Employment

Forschungsschwerpunkt:

Arbeitsmarkt und

Beschiiftigung Abteilung:

Arbeitsmarktpolitik und

Beschaftigung

(2)

ZITIERWEISE/CITATION

Nigel Meager

Self Employment and Labour Market Policy in the European Community

Discussion Paper FS I 93-201

Wissenschaftszentrum Berlin fiir Sozialforschung 1993

Forschungsschwerpunkt:

Arbeitsmarkt und Beschaftigung (FS 1) Abteilung:

Arbeitsmarktpolitik und Beschaftigung

Research Area:

Labour Market and Employment

Research Unit:

Labour Market Policy and Employment Reichpietschufer SO

10785 Berlin

(3)

Abstract

The paper examines recent trends in self-employment (which in many EC countries mark a major break with previous historical patterns), and how those trends are influenced by the economic, structural and regulatory en vironments in the different countries, as well as their various labour market and industrial policy regimes.

The paper falls into two broad parts. The first part begins with a dis cussion about the nature of .self-employment and the theoretical and methodological issues involved in measuring self-employment and comparing self-employment patterns internationally. It goes on to examine the characteristics of the self-employed in EC countries, trends in self-employment, and the dynamics of self-employment (using new flows data from the European Labour Force Surveys, ELFS), and puts forward some hypotheses to explain the different recent experiences of self-em ployment in various EC countries. The paper argues that the factors com monly put forward to explain self-employment trends (including: macro- economic factors, and the "push" effect of growing unemployment; the shift from manufacturing to services; the growth of franchising and con tracting out; demographic change; and attitudinal change) are insufficient to explain the enormous diversity in recent experience in European countries, ranging from a strong growth in the UK at one extreme, to

continued decline in Denmark at the other.

Evidence is presented which suggests that, to explain such diverse de velopments, and to disentangle the influence of policy from the factors mentioned above, it is crucial to take account of the way the different institutional and legislative environments in different countries affect the response of self-employment to economic factors. A comparison of Germany with the UK, for example, shows greater stability of self- employment in Germany, with much lower inflow rates, but cor respondingly lower outflow rates. In crude terms, differences in the in

(4)

Germany than in the UK in 1980s, but having entered self-employment,

German entrepreneurs were better placed to survive in business than their UK counterparts.

The second part of the paper looks in more detail at policies for self-

employment (especially labour market policies), and provides a pre

liminary evaluation of the different policy regimes in EC countries (concentrating in particular on the experiences of Denmark, France, Germany and the UK). The empirical evidence presented suggests that self-employment schemes for the unemployed in the 1980s did make a dif ference, in the sense of increa.sing the flow from unemployment to self-

employment (i.e. they were not purely "deadweight"). It also shows through a detailed comparison of such schemes in four countries, using administrative data, and flows data from the ELFS, that important design features of the programmes (including their eligibility criteria, and whether or not they are allowance-based, or lump sum-based) can have important effects on their impacts (eg in terms of the sectoral composition of the self-employed business, the extent to which existing business are displaced, and the trade-off between deadweight and survival).

Zusammenfassung

Die vorliegende Arbeit untersucht Trends der jiingeren Zeit hin zu mehr Beschaftigung als Selbstandige (was in vielen EG-Landern einen deut- lichen Bruch mitfrilheren Entwicklungsmustern bedeutet). Wie sind diese Trends durch okonomische, strukturelle und regulative Einflusse in den

venschiedenen Landern beeinflusst ? Wie unterscheiden sich unter-

schiedliche arbeitsmarkt- und industriepolitische Ansatze ?

Die Analyse ist in zwei Teile gegliedert. Der erste Teil beginnt mit einer Diskussion Uber die spezifischen Aspekte des Arbeitsmarktsegments der Selbstandigen, iiber theoretische wie methodische Probleme bei der Messung des Anteils der Selbstandigen und bei internationalen Ver- gleichen von Formen der Selbstandigkeit. Sie wird fortgefuhrt mit der Analyse der Gruppe der Selbstandigen in den EG-Landern, der Trends in

(5)

die Selbstandigkeit und der Dynamiken in dem Arbeitsmarktsegment der Selbstandigen. Dabei warden neue Stromdaten der Europaischen Ar- beitskraftestichprobe benutzt. Im Mittelpunkt stehen einige Hypothesen zur Erklarung der unterschiedlichen Erfahrungen der jtingeren Zeit bezuglich der Selbstandigkeit in verschiedenen EG-Landern.

In der Arbeit wird argumentiert, daB die gemeinhin als wichtig er- achteten Einfiufifaktoren zur Erklarung von Trends bin zur Selb standigkeit (einschlieBlich makrodkonomischer Faktoren, des AnstoB- Effektes der wachsenden Arbeitslosigkeit, der Entwicklung weg von der Guter-Produktion bin zum Dienstleistungsgewerbe, der Zunabme ver- scbiedener Formen von Ausgriindungen und von Selbstandigkeit auf der Basis von Francbising-Vertragen, des demograpbiscben Wandels und des Wertewandels) unzureicbend sind, urn den groBen Unterscbied der Entwicklungen in den europaiscben Landern, der von einer starken Zu nabme der Selbstandigen in GroBbritannien bis zu einem bestandigen Riickgang in Danemark reicbt zu erklaren.

Die Ergebnisse legen nabe, daB es fUr eine Erklarung der unter- scbiedlicben Entwicklungen und um den EinfluB des politiscben Handelns

von den oben erwabnten Faktoren zu trennen entscbeidend ist, zur Kennt-

nis zu nebmen, wie unterscbiedlicbe institutionelle und legislative Regel- werke in verscbiedenen Landern die Reaktionen in dem Segment der Selbstandigen auf okonomiscbe Entwicklungen beeinflussen. Ein Ver-

gleicb Deutscblands mit GroBbritannien zeigt beispielsweise eine groBere Stabilitat der Selbstandigkeit in Deutscbland mit sebr viel geringeren Ein-

trittszablen in die Selbstandigkeit und dementsprecbend aucb geringeren

Abgangszablen. Grob gesagt sind in den acbtziger Jabren Unterscbiede im

institutionellen Gefiige gewicbtigere Hindernisse auf dem Weg zur Selb standigkeit in Deutscbland als in GroBbritannien gewesen, aber nacb dem Start als Selbstandige iiberlebten deutscbe Unternebmer zu einem boberen Grad als ibre Kollegen in GroBbritannien.

Der zweite Teil der Arbeit analysiert detaillierter die Politiken (besonders Arbeitsmarktpolitiken) zur Forderung der Selbstandigkeit und bietet eine vorlaufige Evaluation der unterscbiedlicben Politikregimes in EG-Liindern (wobei sicb die Analyse besonders auf Danemark, Frank-

(6)

reich, Deutschland und GroBbritannien konzentriert). Die vorgetragenen empirischen Ergebnisse legen es nahe, daB MaBnahmen zur Forderung der

Selbstandigkeit von Arbeitslosen in den achtziger Jahren einen EinfluB im Sinne einer Zunahme der Falle batten, in denen sich Arbeitslose selb- standig gemacht haben (es handelte sich also nicht ausschlieBlich um Mit- nahmeeffekte). Es laBt sich auBerdem dutch einen genauen Vergleich der entsprechenden Programme auf der Basis von Daten der jeweiligen Ver- waitungen und von Stromdaten der Europaischen Arbeitskraftestichprobe in vier Landern zeigen, daB wichtige Merkmale dieser Programme (einschlieBlich ihrer Auswahlkriterien und ob sie Zuschusse oder Pauschalbetrage gewahren) einen wichtigen Effekt auf die Wirkung haben (bspw. beziiglich der sektoralen Zusammensetzung der Unternehmen der Selbstandigen oder des AusmaBes, in dem bestehende Unternehmen durch neue Selbstandige vom Markt verdrangt wurden sowie hinsichtlich

des Wechselverhaitnisses zwischen ausscheidenden und uberlebenden Un ternehmen).

(7)

Contents

Edited highlights of Report i-xvi

Preface 1

1 INTRODUCTION 4

1.1 Some theoretical considerations 4

1.2 What is Self-employmentAVho are the Self-Employed? 9 1.3 Measurement issues in comparing self-employment: levels or rates? 13

2 SELF-EMPLOYMENT IN THE EUROPEAN COMMUNITY: AN OVERVIEW 15

2.1 Trends in Self-Employment 15

2.2 Trends in individual EC countries 15

2.3 Agricultural and Non-Agricultural Self-Employment 17

2.4 Self-Employment Rates 19

3 A PROFILE OF SELF-EMPLOYMENT IN EC COUNTRIES 24

3.1 The Personal Characteristics of the Self-Employed 25

3.1.1 Gender 25

3.1.2 Age 29

3.1.3 Marital Status 34

3.1.4 Nationality 41

3.2 The Activities of the Self-Employed 43

3.2.1 Industrial Sector 44

3.2.2 The self-employed as employers 50

3.2.3 Hours of work 54

3.2.4 The self-employed and work-related training 58

3.2.5 The self-employed and job search 60

3.3 Regional Variations in Self-Employment 61

4 THE DYNAMICS OF SELF-EMPLOYMENT - INFLOWS AND OUTFLOWS 64

4.1 Inflows and outflows: overall trends 65

4.2 Unemployment and the dynamics of self-employment 69

4.3 Relative inflow and outlfow rates 75

4.4 Composition of inflows and outflows 80

4.4.1 Flows between dependent employment and self-employment 83 4.4.2 Flows between unemployment and self-employment 84 Appendix to Chapter 4: Rates of entry to self-employment by labour market status 92

5 EVALUATING LABOUR MARKET POLICIES FOR SELF-EMPLOYMENT: AN

OVERVIEW 94

5.1 Introduction 94

5.2 Expenditure on Self-Employment Schemes in EC Countries 96

5.3 Scope and Design of Schemes 99

5.4 Comparative Evaluation: Some Research Questions 101

(8)

5.5 Conclusions 107

5.5.1 Problems of evaluation 107

5.5.2 Methodologies for evaluation 108

6 AGGREGATE EVALUATION OF SELF-EMFLOYMENT POLICIES; IS IT

POSSIBLE? Ill

6.1 Introduction: difficulties in evaluating overall scheme impact 111

6.2 Aggregate approaches 113

6.3 Aggregate scheme impact: cross-section analysis 114 6.4 Aggregate scheme impact: time-series analysis - a German illustration 117

7 SELF-EMPLOYMENT PROGRAMMES FOR THE UNEMPLOYED IN THE CASE

STUDY STATES: EVALUATION WITH ELFS DATA 126

7.1 Introduction: self-employment programmes in four countries 126 7.2 Scheme performance: participants' characteristics 127 7.3 Scheme performance: activities of participants 132

8 POSTSCRIPT: CONCLUDING REMARKS 136

REFERENCES 138

(9)

Self-Employment and Labour Market Policy in the European Community Edited highlights of Report

As this report is a relatively long one (the main text being close to 150 pages), we present here in these first few pages a short summary of some of the main points of the report, for the benefit of those readers who are interested in the subject matter, but do not have the time to read the whole report. We do not intend this summary to be a comprehensive one, and we have not included any of the descriptive detail of the profile of the self-employed in EC countries which is to be found in chapters 3 and 4 of the report itself.

Introduction

The report shows that self-employment increased in the 1980s across the EC. The European Labour Force Survey (ELFS) shows that in the nine member states of 1975', between 1975 and 1989, self-employment grew from 12.7 million to 5.5 million and the self-empbyment rate (the share of self-employment in total employment) grew from 12.6% to 14.1%.

Figure 1

EC Seif-employment trends 1973-89 (all sectors)

Total self-employment (1975-100)

200

160

160

140

120

100

80 h

60 I I I I

1873 1875 1877 1879 1881 1883 1885 1887 1888

Source: Euiopeen LabourRtfoe Surveys

Among individual memberstates, however, the picture was extremely varied. Figure 1 shows self-employment trends in four countries (Denmark, France, Germany and the UK) selected for in-depth study in this research^. These trends diverged sharply after 1979, with self- employment growing strongly in the UK at one extreme, and continuing to fall in Denmark

' Namely: B, D, DK, F, I, IRL, L, NL. UK.

' Conducted in 1992 at the WZB under the MISEP programme.

(10)

at the other. France and Germany lie in between, showing slight growth during the 1980s, with some fluctuations. The other EC countries also exhibit considerable variation, although none of them match the extremes of the UK and Denmark (see Chapter 2). The trends occurred against the background of a governmental policy stance broadly and increasingly supportive of self-employment in all EC countries, influenced by the trend towards industrial and labour market deregulation, and the belief that small businesses were an engine of job creation. All EC countries introduced labour market programmes designed to encourage self- employment.

Explaining the trends

Why did EC countries experience such varied trends, given the supportive policy environment,

and increasingly convergent macro-economic conditions? Recent research' examines the

nature and causes of recent self-employment trends, and the role played by labour market policies. A key point emerging from the research, however, is that to evaluate the impact of such policies for self-employment we must recognise that they are but one of a set of inter related influences on the level and composition of self-employment, including:

1 The economic cycle.

Growing unemployment may "push" people to enter self-employment, but equally, small business creation and survival may require a growing economy. These two influences act in opposite directions, so it is not possible to predict the net response of self-employment to changes in the macro-economy. At some times and in some countries "unemployment-push"

dominates, whilst at other times and in other countries, the "prosperity pull" effect dominates.

2 Structural change in the economy.

The shift from manufacturing to service sector employment in EC nations has contributed to growing self-employment, due to the higher rates of self-employment in the service sector.

3 Changes in the organisation and behaviour of employers.

The shift to "contracting out" of service functions, the growth of franchising and similar developments, have contributed to increasing self-employment in some countries.

4 Changing demographic structures.

Different parts of the workforce have different propensities to enter self-employment. Hence factors such as growing female economic activity, an ageing workforce, and a growth in the ethnic minority share of the workforce, result in a changing overall level of self-employment.

' See, in particular. Acs, Audretsch and Evans 1992: Meager, Kaiser and Dietrich 1992, Meager 1992a, 1992b; and OECD 1992, as well as the rest of this report.

(11)

5 Attitudinal change.

It has been argued that there has been a recent change in popular attitudes towards self- employment. In the UK, in particular, the post-1979 Thatcher government's policies, oriented

towards creating an "enterprise culture", are often cited as having led to attitudinal change"*.

The debate on the relative importance of these factors remains open, but recent evidence'

suggests that inter-country differences in these factors are insufficient to explain the divergent self-employment trends. Their combined effects are mediated by a range of influences concerned with the legal and institutional environment, which can help explain differences between countries which might otherwise have rather similar self-employment trends**. Two such influences are: the regulatory framework governing business start-up and occupational entry', and the structure and regulation of the market for finance facing the self-employed.

An Anglo-German example

An Anglo-German comparison suggests that institutional differences result in self-employment in the UK being more "dynamic", with higher entry and exit rates than its German counterpart Despite recent trends in both countries towards deregulation, the regulation of enuy to certain occupations on a self-employed basis, and business start-up in general, remains tighter in Germany than in the UK. This is particularly true in the Handwerk (craft) sector, covering a wide range of activities (from bakers and hairdressers to dispensing opticians), accounting for around 20% of non-agricultural self-employment. To enter self- employment in Handwerk the individual must generally be a Meister (or employ a Meister), which means that h^she must have served an apprenticeship, and have acquired certain post- apprenticeship experience and training. There is no such requirement for comparable activities in the UK. This implies that to a greater extent in Germany than in the UK, self-employment entry in such regulated activities depends on past career choices. Thus, in so far as self- employment entry responds to short-term economic influences, we would expect either that such responsiveness would be less in the more regulated German environment, or that it would be crowded into the least-regulated activities, with low entry barriers.

Further, it was easier for the self-employed to obtain start-up finance in the UK than in Germany in the 1980s. UK financial deregulation led to a credit boom in the mid-late 1980s, with flnancial institutions keen to lend money directly to potential entrepreneurs, or indirectly through loans for consumption or house purchase, which could be easily recycled for business start-up purposes. This contrasts with the tighter credit environment in Germany, and is reinforced in the UK's case by the high and rapidly growing rate of home ownership (much higher than in Germany). This, coupled with rapid house price inflation, led to a growth in personal housing wealth which could be used as collateral for business start-up loans. Capital gains in the housing market often leaked into other areas (e.g. consumption or business start up) through a process of "equity withdrawal". Lastly, there is some evidence that the venture

* Hard evidence of pervasive attitudinal change is, however, scarce (Blanchflower and Oswald 1990).

' SeeAcs et al. (op.cit.) and OECD (op. cit.)

* See Meager et al. (op. cit.). Meager 1992b, and Chapter 1 below.

(12)

capital industry specialising in high risk financing of new businesses is more developed in the UK than in Germany, and expanded during the 1980s.

Such arguments are consistent with evidence from the ELFS of lower inflow rates to self- employment in Germany than the UK, but the new self-employed were also better placed to survive in self-employment in Germany than in the UK, because: (a) more of them would have had relevant qualifications; (b) more would have had their business scrutinised by a lending institution; (c) those entering regulated sectors (like Handwerk) would have enjoyed relatively protected markets. Differences in small business policy reinforce the institutional differences; self-employed Germans face a wide array of central and local government-funded

small business support and advice'. The ELFS shows (see Chapter 4) that German self-

employment entrants are more likely than in the UK to come from wage employment, presumably because the former are better endowed than the unemployed/economically inactive in terms of both financial and human capital. The ELFS also confirms that ouiflow rates from self-employment are lower in Germany than in the UK.

Self-employment dynamics: the need for flows data

This leads to another key theme in the research reported here, namely that explaining self- employment trends requires examination not just of self-employment stocks, but also of entries to and exits from self-employment. Such flows data have been recently constructed for the first time for most EC countries from the ELFS (see Chapter 4), revealing much higher rates of inflow to and outflow from self-employment in some countries (e.g. the UK) than in others (e.g. Germany). They also reveal differences between the sources of self- employment inflows. Thus, in all countries, wage employment is the largest source of new selfemployed, but its relative importance varies considerably, from 31% of the inflow in Spain to 75% in Germany.

The importance of unemployment as a source of self-employment inflows also varies, which is of particular interest, given the debate over unemployment as an influence on self- employment. We show below in Chapter 4, that in the 1980s, with the exception of Germany, the probability of an unemployed person becoming self-employed is in all countries greater than that of an employed or economically inactive person becoming self-employed. Figure 2 shows inflow rates from unemployment to self-employment in the four countries. Most notable is the difference between Germany at one extreme and the UK at the other, and the upward trend in the UK. How can we explain these patterns? Unemployment "push"

hypotheses suggest that as unemployment grows, the unemployed see a worsening in their job finding chances, and feel pressure to consider self-employment. On this view, the inflow rate from unemployment to self-employment would increase with unemployment. The data do not confirm this, however. Indeed in the UK, the inflow rate increased throughout 1982-89, whilst the unemployment rate first increased and then fell strongly. Fuller explanations must, therefore, consider other influences on self-employment inflows, including policies introduced in all EC® countries, aimed at encouraging the unemployed to become self-employed.

' Bannockand Albach 1991 confirm the greateremphasis on support to start-up businesses in the UK as against assistance to existing small flrms in Germany.

* See InforMISEP, No. 21. Spring 1988.

(13)

Figure 2

Entry rate to self-employment from unemployment

% of unemployed entering self-emp during year

j_

198^3 Source; ELFS

I

1983-4

DK

1984-S

UK

_L 1986-7

_L 1987-8

I

1988-9

Labour market policies for self-employment in EC countries OECD data (Table 1) show that such

schemes are a small but growing component of labour market policy expenditure. Their scale varies, with Spain and Greece having the largest schemes, and Denmaric, Germany and Belgium the smallest They typically consist of payments to iliose unemployed who become self-employed, these payments being (wholly or in part) in lieu of unemployment benefits or insurance payments. Hence, even if their net job- creation effect is small, after allowing for deadweight and displacement', the net cost per job created is also likely to be small compared with many other labour market policies. Such schemes effectively involve switching expenditure from "passive" to

"active" labour market policies. Whilst

Table 1: Self-employment schemes for unemployed as % of total labour market policy expenditure

1985 1986 1987 1988 1989

E 1.67 5.61 6.02 6.84

GR 0.00 0.00 0.00 5.05

UK 1.03 1.31 1.88 2.19 2.27

P 0.97 1.89 1.95 1.80

F 1.43 1.49 1.24 1.10 *

IRL 1.49 1.48 1.02 0.74

DK 0.21 0.36 0.41 035

D 0.00 0.06 0.11 0.35 0.15

B 0.00 0.00 038 0.29

EC9 0.88 1.31 1.41 1.61

* indicates data not available Source: OECD

The concepts of "deadweight" and "displacement" are explored further below.

(14)

other active labour market policies (e.g. "make work" or training schemes with an allowance in lieu of benefit), also involve such a switch, these self-employment policies are unusual in that they approach a one-to-one relationship between the outlay on the active measure and the reduction in the passive expenditure on unemployment support.

Evaluation issues

Evaluation of these schemes across the EC raises at least four key issues.

1 Identifying programme objectives

Is the scheme's main objective to reduce unemployment (if so, is this a direct effect or through employing others)? Is business creation a specific aim? Are there less measurable objectives (e.g. impact on labour supply quality)? Most studies do not distinguish the unemployment impact from the self-employment impact, but assume that the two effects are the same. In the short-term, they are equivalent, but over time more complex dynamic effects occun e.g. do participants whose businesses fail, return to unemployment, or get jobs because being self-employed has made them more employable?

Our interest is in the self-employment impact, but to explain differences between scheme effects we must consider (often subtle) differences in scheme objectives. For example, the

UK's Enterprise Allowance Scheme (EAS) and its German equivalent, Oberbriickungsgeld,

seem similar in design and objectives. Both pay an allowance to unemployed entrepreneurs, with wide eligibility including the short-term unemployed. For both, unemployment reduction is a key objective, although the German scheme is set in a more traditional labour market policy framework, whilst the UK scheme is also concerned with promoting the "enterprise culture". Further, a clear objective of the German scheme is the Vermeidung von unterwertiger Beschliftigiing (avoidance of undervalued employment), and applicants must show that their self-employment generates a specified minimum income. The official view in the UK is different; studies show that EAS recipients have very low earnings in self- employment, but these are seen as a supply side benefit of the EAS, through reduced wage pressure in the economy'". In 1988, 80% of EAS recipients earned less than the monthly

3,400DM threshold for Oberbriickungsgeld. The EAS was often criticised" for encouraging

entry to low margin, highly competitive activities, which have poor survival chances, or displace existing businesses. On this interpretation, however, such effects can be seen as consistent with the scheme's stated objectives.

2 Deadweight

Deadweight is a risk in such schemes, in that many participants may have entered self- employment anyway (without a subsidy). Short-term deadweight on the UK scheme, for example, may be as high as 40%'^, although doubt has been cast on such estimates based on retrospective surveys of participants. Nevertheless, surveys in other countries have

See Owens 1989 (page 6, in particular).

" See Gray and Stanworth 1988.

See PA Cambridge Economic Consultants, 1990.

(15)

suggested lower deadweight; only 17% of participants in the Danish scheme, for example, would definitely have become self-employed without the scheme". The larger the deadweight, the smaller the scheme's net effect, and a key evaluation issue, therefore, is

which schemes have higher deadweight, and why.

3 Displacement

Even when participants would not have entered self-employment without the subsidy, we can ask if their businesses affect existing (unsubsidised) businesses, by taking market share or by driving them out of business. As with deadweight, the larger the displacement, the smaller the scheme's net impact. Calculating net impact is difficult; it depends on what happens to those who are displaced (do they enter unemployment, obtain wage employment, or leave the workforce?). Further, it is possible that displaced businesses would have survived longer than those displacing them. Displacement is harder to estimate than deadweight, and surveys of scheme participants are unreliable here'"'. In the UK, displacement has been assessed at 50%,

but this is a "guesstimate", and in the only detailed study of displacement", based on

exhaustive study of a local labour market, estimates of close to 100% displacement were obtained in the hairdressing sector. Displacement varies with the sector entered by participants; in particular, it tends to be higher in crowded markets with low entry barriers.

4 Differences in scheme design and implementation

Scheme performance depends not just on the external environment into which the scheme is introduced, but also on a number of important scheme design factors including:

a Eligibility

The schemes are designed to increase the flow from unemployment to self-employment, so eligibility is normally confined to the unemployed, in many cases covering anyone entitled to unemploymentcompensation. Some countries (e.g. Germany, the UK and Ireland) exclude

very short-term unemployed, whilst two countries (Portugal and Denmark") confine

eligibility to the long-term unemployed (Denmark also imposes an age criterion). Broad eligibility does not mean broad participation, however, and studies from several countries show that even where all unemployed are eligible, scheme participants are concentrated amongst the more advantaged unemployed (better qualified, male, with shorter durations of unemployment). A priori, it is not clear which kind of scheme design is more effective.

Schemes with wide eligibility run a risk of high deadweight, whilston the other hand schemes targeted at disadvantaged groups may have lower deadweight, but poorer survival rates.

" See Rosdahl and Maerkedahl 1987.

Elias and Whitfield 1987, discuss the methodological difficulties of measuring displacement

Hasluck 1990

In Denmailc, eligibility was widened to include short-temi unemployed in 1989.

16

(16)

b Whether the subsidy is paid as an allowance or a lump sum

The schemes also differ in the way participants are paid. Some (e.g. France and Spain) pay a lump-sum grant in advance (in Spain this is a capitalisation of the benefits he/she would have received whilst unemployed). Most pay an allowance, similar to the payment of unemployment compensation. There is little difference between the two from an exchequer viewpoint, but they may have different outcomes in terms of labour market dynamics. In particular, the funding mode may affect the types of participants (a grant might, for example, be better at attracting those who would not otherwise enter self-employment), and the types of activity entered, which may affect influence survival rates. We noted above that a criticism of the UK scheme, (an allowance) was that it encourages entry to markets with low entry- barriers, and low returns. In these largely service sector markets, survival rates are low and displacement high. A key question, therefore, is whether a grant-based subsidy helps to avoid this problem and achieve a "better" overall sectoral distribution of scheme participants.

c SuDDort and training

The existence of appropriate support and training for scheme participants may weaken the trade-off between deadweight and survival identified above.

d Scale and duration of pavments

Here one needs to consider the inter-related questions of the size of the payments to scheme participants, the duration ofpayments, and the number ofparticipants. Is it more effective to spend a given budget for such schemes on a small number of participants receiving a relatively large payment for a long time, or to spread the money more thinly, with more participants, smaller payments and shorter durations of payment?

Methodologies for evaluation

What empirical methodologies exist for investigating these evaluation issues, particularly in a cross-country context? Ideally, evaluation would involve control-group experimental approaches, as in the USA". Such approaches have, however, been ruled out in most EC countries on grounds of cost, ethical objections to experimental methods, or because political expediency required rapid scheme implementation. Evaluation must, therefore, be conducted ex post, and there have been three empirical strategies for evaluating these schemes:

1 Scheme administrative data/follow up surveys

The commonest strategy, adopted in most countries, is to use administrative data to identify the characteristics of scheme participants, together with surveys to identify their perceptions, experiences, survival rates etc. Its main drawback, apart from cost and time, is that without comparisons of non-scheme entrants to self-employment, strong conclusions cannot be drawn.

Equally, reliance on participants' perceptions/recollections to identify deadweight, and especially displacement, is fraught with difficulties (although these can be eased by seeking

" Sec Wandner and Messenger 1991

(17)

collaborative perceptions from competitors and customers of scheme participants; see Elias and Whitfield 1987). For internationally comparative evaluation, we would argue that there is some scope for using such data in conjunction with base data from the ELFS (see below).

2 Aggregate impact studies

Another approach is to identify the effects of the schemes in aggregate data, controlling, if possible, for other economic and environmental influences. This analysis can be conducted on a cross-section basis (examining scheme impact at a point of time across countries) or through time-series analysis Cooking at scheme impact over time within a single country).

Figure 3

Self-employment schemes and Inflows from unemployment (1987-8)

% Inflow rato to self-employment from unemployment (1087-8) 5n

4.5- 4 3.5-

3- 2.5

2 1.5

1 •]

0.5

o p

IRL

° DK

B

0 o!5 ] iY

Source: ELFS and OECD1990

E F

2.5

UK

GR

~5 sli 5 4!5

Entrants to sclieme In 1088 as K of 1087 unemployed

Cross-section analysis looks for a relationship between the number entering self-employment from unemployment in a country, and the number of participants in that country's scheme.

If deadweight is high, we would not observe such a relationship, since the scheme would make little difference to the flow from unemployment to self-employment. Figure 3 plots these data for 1988 in nine EC countries, showing a clear positive relationship. Countries whose schemes have higher participation rates have higher rates of entry from unemployment to self-employment. Inter-country variations in scheme intensity explain a significant part of variations in the inflow to self-employment from unemployment. This clear relationship suggests at the very least that the schemes have some effect, tending to increase the inflow

(18)

from unemployment to self-employment by a non-negligible amount; that is, deadweight is

less than 100%.

There are few examples of time-series studies with self-employment schemes as explanatory variables. An exception is Johnson et. al. 1988, whose results provided only weak evidence for a positive effect of the Enterprise Allowance Scheme on UK self-employment. Such studies, however, are based on an analysis of self-employment stocks. As a key objective of the schemes is to increase the inflow from unemployment to self-employment, we need also to examine the effect (net of deadweight) of the schemes on inflows to self-employment.

Similarly the combined effects of non-survival of scheme participants and displacement of non-participants can be estimated only by modelling the outflows from self-employment.

Unfortunately ELFS flows data are not available for a long enough period for time-series modelling. A few countries do, however, have appropriate time-series data, and we present

in Chapter 6 below, a German example'* showing that the German scheme, although smaller

than those in some other EC countries, appears to have a significant impact on the rate of flow from unemployment to self-employment, and that the scheme may have relatively low deadweight

3 Inference through comparisons with representative micro-level data

In comparing self-employment schemes, however, some use can be made of ELFS flows data to look at the characteristics of the newly self-employed, of those entering self-employment from unemployment and of outflows from self-employment. Comparisons of these patterns with the characteristics of participants may then provide insights on scheme performance. This approach is adopted in some depth in the research reported here.

To give a more concrete illustration, some working hypotheses might be as follows. Firstly, the more a scheme is targeted (at people with characteristics not normally found in the self- employed), the less will be initial deadweight, but the less will be survival rates (unless accompanied by effective support and training). The ELFS can provide a profile of the personal characteristics of entrants to self-employment in each country. By comparing this profile with the characteristics of scheme participants, we can make a preliminary assessment of whether deadweight is likely to be high or low. Similarly, by examining outflow rates of people with "atypical" characteristics for self-employment, it may also be possible to provide indications of the likely increased failure rates associated with targeted schemes.

Secondly, grant-based schemes are more likely to encourage entry to markets with high entry barriers but good income and survival prospects, whilst allowance-based schemes encourage entry to markets with low entry barriers and low capital requirements. Displacement is likely to be smaller in the first type of market, as new entrants reduce the (protected) profits of existing businesses rather than drive them out of business. The ELFS data can indicate the business activities entered by the newly self-employed, for comparison with the activities of scheme participants in different countries, to assess whether the schemes shift the profile, and whether grant-based schemes shift it in a different way from allowance-based schemes. The ELFS data enable us to compare scheme performance taking account of the underlying

" Using now data constructed from the Arbeitskrafie-Gesamtrechnmg.

(19)

structure of self-employment flows in each country. Thus if, for example, the proportion of scheme participants in manufacturing is higher in Germany than France, this need not imply that the German scheme is more successful than the French in encouraging the self-employed to enter manufacturing, since it also depends on the proportions of the overall flow from unemployment to self-employment entering manufacturing in each country.

Table 2: Summary of self-employment schemes for the unemployed in four EC member states

Germany France UK Denmark

Name of scheme OberbrUchtngsgeU Aide awe chimeurs erdttteurs d'entnprises

Enterprise Allowance Scheme Ivarksalteryddsen

l>ate introduced 1986 1979/80 1983 1983

Eligibility Registered unemployed (after 11 weeks; reduced to 4 weeks in 1988). No legal entitlement (Bundaanstalt fiir Arb<il can operate budgetary ceiling for scheme).

All receiving or entitled to unemployment benefits (recently extended to include recipients of other welfare benefits).

Unemployed for at least 8 weeks and receiving unemploymentAiupplemeotary benefit (family credit). From 1991 limit reduced to 6 weeks, and local Training and Enterprise Councils (TECs) have discretion on rules.

Long-term unemployed over 23 (unemployed at least 21 months and had 1 statutory 'job offer"). Revised 1989 to include LTU under 23 (12 months unemployment and 1 'job offer'), and others with at least 3 months uaertiploymcnt.

Form of support Monthly allowance Capital grant Weekly allowance Monthly allowaace Rote of payment Equivalent to previous

beneflt enUtlement (ceilings introduced and progressively reduced

scheme revision in

1988), plus contributions to social security costs.

Between FF. 10,730 A 43,000, (falls with length of unemployment); plus exemptions from sorrw social security payments.

(Extra grant if new enterprise creates Jobs).

Rat rate of £40 per week, from 1991. individual TECs have discretion to vary payrtKnts (from £20-£90 per week)

30% of maximum

unemployment benefit (up to a ceiling of DKr.34,000 pa.)

Duration of paymenis

Up to 3 months. Up to 6 titonths from 1988; and duration of payments increase with duration of unemployment.

One-off payment Up to 32 weeks. From 1991 TECs have discretion to vary payment period (from 26 to 66 weeks)

Up to 3.3 years

Conditions for receipt of payment

Must have proposal approved by competent authority (bank, chamber of comttierce, professional assoc. etc);

business must guarantee likely minimum income of DM3,400 p.tn

No restrictions initially.

Since 1987. applicants mutt fill in detailed questionnaire, and are vetted on the likely viability of their proposed

business

Must hove £13)00 own capital.

Must work for at least 36

hours per week. New business

imrst be the whole InvestmenL Some 'unsuitable* business activities excluded. From 1991 TECs can vary criteria. e.g. to require approved business plan.

No specific restrictions on applicants, or type of trusiness. No own capital requirements.

Support mechanisms None tied to scheme, but participants eligible for wide range of support through chambers of

commerce etc.

None tied to scheme, but participants eligible for normal state-fimded business advice etc

'Awareness day* prior to enlry. Participants visited at least once by officials in year.

Option of 3 free business counselling sessions in year.

Recipients have option of participating in special courses for new

entrepreneurs at technical or commercial schools

Number of participants

1986; S.728 1987: 9.996 1988: 173183 1989: 11,242

1979: 9,200 1986: 71.377 (peak) 1990: 49.316

1983/4: 27,600

1937/8: 106J00 (peak) 199(V1:

60,300

1983; 409 1987: 1,008 1989: 3,308 1990: 3,641 Survival rates Not available, but of

scheme participants in 1987. some 8% were unengiloyed by May 1988. Bgure for 1988 participants was 6%.

After 1 year about 83%

still trading. Alter 2 years 73%. Highest survival in construction and

manufacturing, lowest in

commercial services.

Of those who complete 12 months on the scheme, about 3/4 survive a fhrther 6 months, and about 2/3 survive a further 2 years.

About 76% of starters in 1989 were in business 2 years later. Earlier studies show that al>out 29% of participants do net surtdve the 3.3 years of eligibility.

(20)

We use, in the research reported here, ELFS data to compare self-employment schemes in detail in four countries (Germany, France, the UK and Denmark). Table 2 summarises the schemes' main features. The key differences are the relatively large scale of the UK and French schemes; the fact that the French scheme is a capital grant, whilst all the others are allowance-based subsidies; the initial targeting of the Danish scheme on the long-term unemployed (later widened); the wide eligibility of the UK scheme, and (until 1987) the French scheme, whilst the German scheme requires greater scrutiny of proposed businesses.

Of the three allowance-based schemes, the UK allowance is the smallest (the maximum an individual could receive under the scheme was a total of 3,016 ECU in 1988) and the Danish the largest (the maximum receipt was 23,800 ECU in 1988).

Figure 4

% female

35

25

15

Self-employment programmes gender oompcsltion (UIQ

Enterprise Allowanoe

Scheme

Aggregate How: unemployment to self-employment

19^5

1986-7

19^-

Figures 4-7 compare the gender distribution of scheme participants with that of the aggregate inflow to self-employmentfrom unemployment in the four countries. The UK scheme shows strong growth in female participation (from 15% in 1983-84, to 36% in 1989). This partly results from deliberate policy, following a perception that women were initially under-

represented on the scheme*'. This development, whilst apparently successful in targeting a

group disadvantaged in terms of access to business start-up, and therefore reducing likely deadweight, was not at the expense of survival rates (EAS evaluation studies show that survival rates of women were lower than those of men - partly because they tended to set up different types of business - but overall survival rates rose slightly over the period as a result of the improving UK economic situation in the late 1980s).

" A conclusion drawn by comparison with women's share of the stock of self-employment.

(21)

Figure 5

Self-employment programmes

% female gender composition PK)

Ivaerksaetterydelsen ^ " \

^ \

^ ^ \

\

V N

\ V

65

60

55

50

45

40

35

30,

Aggregate flow: unemploymentto self-employment

L I _ 1 ^

i m 5 i9P3 1986-7 1988-9

Women are also heavily over-represented in the Danish scheme compared with the overall inflow to self-employment. This results from the high concentration of women in the long- term unemployed target group, and the proportion of women in the scheme fell sharply when the eligibility criteria were widened in 1989. Women are more under-represented among the self-employed in Denmark than in other countries (only 15% of Danish self-employed are women), and it seems that the targeting of the Danish scheme on disadvantaged groups involved relatively low deadweight. This is confirmed by early surveysof scheme participants showing that only 17% would definitely have set up in self-employment without the scheme (a much lower proportion than in UK surveys of HAS participants, for example).

The Danish scheme appears to avoid the expected low survival rate of a targeted, low- deadweight scheme, by combining a generous allowance (for up to 3.5 years) with extensive support and training for participants.

In France the proportion of female participants fluctuates but is broadly in line with the female share of unemployed entrants to self-employment. The German scheme stands out, however, with a clear under-representation of women, reflecting its more stringent entry criteria: a businesses viability must be externally assessed, and it must generate a specified income for the participant This latter criterion, in particular, may discriminate against female applicants; UK studies show that female EAS participants arc more likely than males to enter low margin, service activities such as hairdressing and similar services. Many such activities are excluded from the German scheme by the income requirements.

Our discussions above suggest that such restrictions would result in high deadweightand high survival rates. In practice, the German data do not suggest higher than average deadweight,

(22)

Figure 6

% female

40

30

20

1982:3"

Figure 7

% female

30

26

22

18

Self-employmentprogrammes gender composition (D)

1984-S

Aggregate flow: unemployment to self-employment

Uebertmjeckungsgeld

1986-7

Self-employment programmes gender composition (F)

Aggregate flow: unemployment to self-employment

Chomeurs Createurs

-9

i m r i m r 1988:9

although the post-scheme re-entry rate to unemployment (about 6-8%) is consistent with high survival rates. Low deadweight may result from the tapering of payment (the subsidy increases with previous unemployment duration). If the unemployed who enter self-

(23)

employment without a subsidy tend to be short-term unemployed, this tapering may help reduce deadweight. The scheme design, with greater weight to the long-term unemployed, and strict criteria on business viability and income potential, may increase survival rates without increasing deadweight unduly, but at the expense of reducing female participation.

Figure 8

Self-employment programmes sectoral composition 1987-80

% in production industries

20

IS

10

] Aagreoate flow

unemp to seff-emp

Scheme participants

Further analysis (Chapter 7) examines the activities of scheme participants in the four countries, compared with those of the newly self-employed who were previously unemployed.

The UK scheme (HAS) has an over-representation of small scale service activities (consistent with the hypothesis that allowance-based schemes may emphasise activities with few entry barriers and financial or human capital requirements). Manufacturing is, however, over- represented among participants in the German scheme, compared with the aggregate flow to self-employment from unemployment (Figure 8). The ELFS data show that manufacturing accounts for a higher share of self-employment in Germany than in other EC countries, but that the unemployed are disadvantaged in entering the sector (in line with the German institutional features discussed earlier). It seems that Oberbriickungsgeld may help redress that imbalance, with its relatively generous support, and strict scrutiny of eligible businesses.

Manufacturing and Handwerk activities are, for example, more likely to satisfy the income criteriaof Oberbriickungsgeld, than are service activities {Gout and Biichtemann 1987confirm that average self-employed incomes are much lower than those in manufacturing).

Finally, the French data are consistent with the notion that grant-based schemes may shift the composition of supported businesses away from low-earning, service sector activities, by helping to overcome entry barriers to capital-intensive activities. French scheme participants (Figure 8) are heavily over-represented in production activities.

(24)

Concluding remarks

To conclude, we stress the importance of inter-country differences in the institutional environment in explaining recent self-employment trends in EC countries. The ELFS flows data reinforce this argument, as well as providing a new dimension to comparative evaluation of self-employment schemes for the unemployed. The ELFS analysis suggests that the existence of such schemes does make a difference: they are not purely "deadweight"; more unemployed people enter self-employment due to these schemes than would otherwise be the case. This does not, of course, imply that the self-employment generated is sustainable (UK experience in the current recession is stark evidence of this). Neither does it imply that all scheme designs are equally effective; our evidence suggests that selective eligibility criteria and/or a grant-based rather than allowance-based payment may help to shift such schemes towards the development of sustainable self-employment with significant earnings potential (and perhaps subsequent job generation potential).

(25)

PREFACE

This report, as the title implies, is about self-employment in EC countries. More specifically it is about the recent trends in self-employment (which in some EC countries at least, mark a major break with previous historical pattems), and how those trends are influenced by the economic, structural and regulatory environments in the different countries, as well as the various labour market and industrial policy regimes in those countries.

The 1980s saw a notable increase in self-employment across EC countries, both in absolute terms and as a proportion of total employment. Within this overall picture, however, the experience of individual countries was highly variable, with the UK experiencing by far the greatest increase, in contrast to others (such as France and Germany) where the level (and rate) of self-employment was stagnant, and still others (such as Denmark and Luxembourg) where it continued to decline. These varied trends, moreover, occurred against a policy background which was broadly and increasingly supportive of self-employment in all EC countries. Without exception, since the late 1970s, member states introduced a variety of labour market programmes designed to encourage members of the labour force (particularly unemployed members) to enter self-employment, and to support those who did so.

It is therefore timely to attempt to assess the significance of some of these recent developments. Key issues concern:

• the extent to which apparent differences between EC countries' experiences are also

real ones;

• the causes of the differential trends in self-employment between the different

countries; and

• the disentangling of the effects of specific policy measures from the influence of other changes (economic, social and structural) occuring in the countries concerned, on both the demand and supply sides of the labour market.

• whether the apparent reversal in many countries of the earlier trend towards declining self-employment, represents the re-emergence of self-employment as a new and significant form of employment in the "post-Fordist" era of industrial organisation {Loveman and Sengenberger 1990), or whether the experience of the last two decades is simply a temporary deviation from the downward historical trend (associated with particular economic and social circumstances);

The research reported here begins to address these complex questions, using comparative data on labourmarket developments (from the European Labour ForceSurveys), together with the findings of existing research on self-employment in the individual countries, as well as administrative and other information on relevant policy initiatives in the different countries.

Much of the general analysis of trends in and patterns of self-employment presented in the

(26)

report covers all EC member states', whilst the more detailed analysis of policies and institutions concentrates on the experience of four selected member states (Denmark, France, Germany^ and the United Kingdom).

The research was undertaken while the author was a Senior Guest Fellow at the

Wissenschaftszentrum Berlin fUr Sozialforschung (WZB), and forms part of the series of evaluation studies conducted by the WZB for the Commission of the European Communities

(DG V), under the MISEP' programme. It also draws in part on earlier research on self-

employment in the United Kingdom and in Germany, conducted by the author (see Meager 1991 and Meager, Kaiser and Dietrich 1992). The report itself is a summary of the work conducted during the course of the project Several more detailed papers and articles on different aspects of the project were also produced, including "issues papers" (on unemployment and self-employment, taxonomies of policy regimes, and evaluation methodologies), as well as literature reviews pertaining to self-employment in individual member states. Some of these have been published elsewhere and are referred to in the text below; others have not, but are available from the author on request.

The author is grateful to the WZB Research AreaArbeitsmarktpolitik und Beschaftigmg, and its director Prof. Gunther Schmid for the invitation which provided the opportunity to undertake this research. He is grateful also for the capable research assistance provided at various stages of the work by Antoine Deberdt, Christel Degen, Herrad Hdcker and Lothar Linke, for secretarial support from Karin Reinsch, and to other colleagues for suggestions and ideas during the course of the work (notably Peter Auer, Christoph Biichtemann, Hugh Mosley and Bemd Reissert). No blame attaches to any of these people, however, for any errors or inadequacies in the report.

The report is structured in two parts as follows:

Part 1 (chapters 1-5) provides general background about the extent and nature of self- employment in EC countries, examines trends in self-employment, and the dynamics of self- employment, and puts forward some hypotheses to explain the different recent experiences of self-employment in various EC countries. In particular:

• Chapter 1 introduces the topic with a discussion about the nature of self-empbyment and the theoretical and methodological issues involved in measuring self-employment and comparing self-employment pattems internationally;

• Chapter 2 provides an overview of recent trends in self-employment in EC countries;

' The main exception here is that in much of the more detaiied analysis, we exclude Luxembourg, since the combined effects of looking at a minority population (the self-employed), through a relatively small sample (the ELFS) in a vexy small country (Luxembourg), are that data cell sizes are typically too small, due to sampling error, for statistically reliable conclusions to be drawn.

^ Throughout the report "Germany" refers to pre-unification (West) Germany.

^ Mutual Information System on Employment Policies

(27)

• Chapter 3 constructs (using data from the European Labour Force Surveys) a cross-

sectional profile of the self-employed in the EC, identifying their key personal characteristics, and the characteristics of the activities in which they are engaged;

• Chapter 4 looks at the dynamics of self-employment, again using ELFS data to constructa pictureof "flows" between self-employment and other labour marketstates;

Part 2 (chapters 5-7), looks in more detail at policies for self-employment (especially labour market policies), and provides a preliminary evaluation of the different policy regimes in EC countries (concentrating in particular on the experiences of Denmark, France, Germany and the UK).

(28)

INTRODUCTION

1.1 Some theoretical considerations

In addressing questions about the nature and causes of recent self-employment trends in Europe, the literature of theoretical social science does not provide us with a Arm base from which to begin our enquiries. Self-employment has not, until recent years, attracted signiAcant attention from social scientists. As other recent authors have noted (e.g. Steinmetz and Wright 1989), since the 1950s labour market sociologists had either ignored self-employment, or treated it as a residual segment of the labour market, associated with agriculture, small scale crafts, and a somewhat archaic "petit-bourgeois" section of society. Whilst there is a strand of sociological literature examining the characteristics of the self-employed as individuals and as members of a distinct social stratum (see Burrows 1991), beyond the notion that this stratum is a declining one there has been little aggregate analysis (for recent exceptions see Steinmetz and Wright 1989, Bogenhold and Staber 1990). Thus, whilst a fair amount is known about the social and economic characteristics influencing individual propensities to become self-employed, these analyses do not typically provide us with a basis for explaining macro-

level trends.

The treatment of self-employment in the labour economics literature has been even more limited, and again this seems in part to be due to its having been seen as an obsolescent mode of employment, largely characterising pre-industrial societies. Indeed, most modem labour economics texts provide no more than passing reference to self-employment. To take one, by no means unique example, we may refer to Ashenfelter and Layard's authoritative and influential labour economics handbook, which aims to bring together,

" a systematic review of the research topics, empirical Andings, and methods that comprise modem labour economics" {Ashenfelter and Layard 1986, Preface).

The sole reference to self-employment as a topic in this two volume, 1300 page work^, is a

brief discussion at the beginning of a chapter on the "employment relationship", in which the author comments on the historical decline of self-employment in the US to a low point in the 1970s, and its (implicitly) inevitable replacement by larger organisations, structured around the wage employment relationship. As the author notes, this historical development has conditioned the predominant emphasis on dependent employment in the labour economics literature (see Parsons 1986, p 791.).

One possible reason for the neglect of self-employment, at least by neoclassical economists, is that as a topic it falls rather uneasily between two sub-branches of the discipline. Thus it is often unclear whether it is rightfully within the purview of labour economics or whether

A more recent British review of the area {Sapsford and Tzannatos, 1990), which claims to provide an overview of the "frontiers" of the subject for undergraduates, contains no references to self- employment at all.

(29)

it is better treated within the framework of industrial economics. From a labour economics

perspective (see, for example, Rees and Shah 1986, Blau 1987), it is common to model entry into self-employment as a problem of labour supply or occupational choice, with individuals maximising their utility across various combinations of income and leisure, and the various

"non-leisure" options including both dependent wage employment (as in traditional neoclassical labour economics), and self-employment. A key problem inherent in such approaches is that it is implausible to treat self-employment as just another form of employment, with the income received by the self-employed regarded purely as a return to labour input, in much the same way that the wage is regarded as a return to labour input in dependent employment. Introducing more realism into such models typically makes for greater complexity and less tractability, since it is clear that income in self-employment also usually includes some elements of a return to capital. Hence the way in which this should be treated may depend importantly on whether the self-employed person owns the capital in question him/herself, or whether the latter is supplied through the capital market (in which case the whole question of the degree of perfection of capital markets facing potential entrepreneurs - highly relevant to policy evaluation - needs also to be considered). The model may also need to be broadened to allow for the fact that returns to self-employment are also in part a return to risk-bearing.

Compared with the labour economics tradition, there is a more significant literature on the subject in industrial economics. Much of this latter work, however, concentrates on the creation and deaths of firms, barriers to entry, limit pricing etc. It typically fails, moreover, to allow for the fact that self-employment is partly (peihaps largely) a labour market phenomenon, and that the self-employed may not always be best characterised analytically as "entrepreneurs", or simply as very small businesses (although clearly many of the self- employed are small business owner/proprietors). The mainstream models of industrial economics and the "theory of the firm" suggest that due to economies of scale and technological requirements, the optimal size of an "enterprise" invariably implies an employment level greater than one. Thus the self-employed are Ueated as an (extreme) case of "small enterprises", and self-employment is of interest primarily as a transitory stage through which (some) newly emergententerprises might pass on their way to becoming fully- fledged "small businesses".

Some of the more interesting examples of the recent economic literature on this subject, however, have begun to develop models of "entrepreneurship" {Evans and Jovanovic 1989, Blanchflower and Oswald 1990a), which integrate many of the factors discussed above at a micro-theoretical level, and which take some steps to bridge the theoretical gap between labour economics and the theory of the firm. Again however, these approaches often still lack a clear exposition of self-employment as a labour market phenomenon, and do not allow for the possibility that a significant proportion of the self-employed may not be "entrepreneurs"

in the usually understood sense of the word, but may be highly dependent on larger organisations.

In any event, the linkage between such theories and any aggregate model remains rather weak, and they provide a poor starting point for an attempt to explain observed trends in self- employent at a macro level. It would seem, therefore, that from the viewpoint of the research exercise reported here, there is no single coherent existing theoretical framework which can

Referenzen

ÄHNLICHE DOKUMENTE

business. By allowing workers to get affordable insurance through the individual market the ACA will substantially increase the rate of small business formation. 2) There is

In particular, the self-employed (employees) have a comparative advantage in their chosen employment status if the estimated coefficient for the selection effect in the

While growth itself is undoubtedly responsible for returning vast swathes of the unemployed back to the labour force, the sustained fall in long-term unemployment is equally

The crucial aim of our paper is to investigate, in a generational perspective, the dissimilar patterns in employment status across a set of four transition

Almost 65 percent of the self employment based households are associated with microfinance program, while about 48 percent of the wage income based household keeps link

The variable measuring the proportion of people in agriculture is positive and strongly significant, and suggests that a 1% increase in the share of people in agriculture increases

While basic organizational principles underlying the new supplier network concepts have been maintained, the relationships between final assemblers and suppliers within both regions

The central message that this paper conveys is that the deep-seated economic institutional configurations that comprise national production regimes provide a crucial clue