the impact of welfare and labour market institutions on informal ...

European Societies 8(4) 2006: 555 /576 – 2006 Taylor & Francis ISSN 1461-6696 print 1469-8307 online

THE IMPACT OF WELFARE AND LABOUR MARKET INSTITUTIONS ON INFORMAL RECRUITMENT IN EUROPEAN YOUTH LABOUR MARKETS Ivan Harsløf Oslo University College, Research Group for Inclusive Social Welfare Policies, P.B. 4, St. Olavs plass, NO-0130 Oslo, Norway

ABSTRACT: Youth labour markets where informal recruitment practices predominate are likely to amplify existing inequalities among young job seekers. Whereas most literature on informal recruitment focuses on characteristics of individual job seekers and the nature of the jobs they obtain, this article suggests to relate this important issue to overarching dynamics of welfare and labour market institutions. Analyses of survey data among young people with a history of longer-term unemployment in eight European countries suggest that comprehensive welfare state arrangements may substitute for the importance of personal network resources in the job search process. Thus the welfare state may intervene by providing active measures to facilitate the job-matching process and by providing economic means to make young people less dependent on their social network. The level of youth unemployment also seems to be related to the extent of informal recruitment, which is found to be more widespread in the countries with high rates of youth unemployment. Key words: active labour market policies; employment protection legislation; informal recruitment; social capital; unemployment compensation for young people; youth unemployment

1. Introduction

For young unemployed persons with limited or no work record to refer to, personal contacts may be of great importance for their chances of gaining a foothold in the labour market. As has been pointed out by Granovetter, groups with a marginal position in the labour market often depend on strong ties to friends and family for lack of professional contacts in the job search process (Granovetter 1995: 148 9). However, labour markets, in /

DOI: 10.1080/14616690601002616

555

EUROPEAN SOCIETIES

which jobs are largely allocated on the basis of the personal network resources that jobseekers can mobilize, are likely to amplify differences originating in the unequal possession of other resources (Barbieri et al. 2000: 217; Lin 2001: 64 5, 166; Portes 1998: 14). Among other things, it has been pointed out that informal recruitment may perpetuate gender inequalities. Thus, women have been found to have a ‘return deficit’ compared to men when utilizing their social network in job search processes (Lin 2000). In general, social networks have been found to be ‘of general importance for social stratification’ (Korpi 2001: 168). For these reasons, lessening the importance of personal network resources in the job allocation processes when young people are to enter the labour market can be regarded as a critical task for the egalitarian welfare state. The characteristics of individual job seekers and the nature of the jobs they occupy have achieved most focus in the litterature on recruitment patterns (Granovetter 1995: 148 9). Research on the linkage between informal recruitment and national welfare policies and labour market policies is scarce. The immense literature on active labour market policies has by and large ignored the importance of this type of labour market intervention on how jobs are allocated (for an exception see Raffe, 1985, who has studied the importance of the Scottish public career service and employment programmes on informal recruitment). Furthermore, there is a lack of research on recruitment processes in comparative perspectives. A comparative perspective may offer important insights to the understanding of the importance of national policies and structures in the labour market. The objective of this article is to explore institutional determinants of informal recruitment pertaining to young people on the margin of European youth labour markets. With reference to the literature on social capital it is argued that structural features originating in different welfare and labour market policies should be considered in order to understand the phenomenon of informal recruitment. The article investigates different perspectives on informal recruitment, examining the importance of ‘active’ polices (Public Employment Service, activation measures), and ‘passive’ policies (income replacement), the strictness of employment protection legislation and the level of youth unemployment. To explore the importance of such macro institutional features, survey data from different European countries are employed. Samples of young people with a history of longer-term unemployment from Finland, Norway, Sweden, Denmark, Scotland, Germany, Spain, and Italy are included in the study.1 /

/

1. Iceland and France were also included in the data but are excluded from the present study: Iceland because comparable macro institutional variables are not available and France because the French study did not include information necessary to create the ISCO 88 occupational category variable pertaining to the reference job.

556

Informal recruitment in European youth labour markets

HARSLØF

In section 2, theoretical perspectives on welfare arrangements and labour market institutions in relation to recruitment patterns are discussed. In section 3, data and methods are presented. Section 4 compares macro institutional features of the countries considered, relating to these different perspectives. In section 5, results are presented. Section 6 concludes the article by discussing the results.

2. Theoretical perspectives on informal recruitment

In the sociological approach to the study of how jobs are allocated in the labour market, a primary assumption is that labour markets are embedded in a more subtle social structure. Hence, individual success or failure in the labour market is not only to be explained by individual propensities such as skills and motivation, or alternatively by characteristics of the socio-economic setting in which the individual finds herself, e.g., in terms of labour demand in the local labour market. In addition, individuals are highly dependent on the labour market success and failures of those belonging to their accumulated reservoir of personal contacts (Granovetter 1985; Granovetter 1988: 191 7). This sociological perspective on the job matching process could be argued as having special importance when it comes to youth labour markets. This is due to the important role of the family of origin in this phase of life. Studies have shown that the family is a decisive alternative welfare provider for young unemployed persons, the extent, however, varying with the type of welfare regime (Bison and Esping-Andersen 2000).2 In circumstances of strong family interdependence the pooling of social network resources is encouraged, as all members of the household may have an interest in ensuring that other household members are employed. The sociological emphasis on social network relations has been argued to place this theoretical approach in between neo-classical economic approaches focussing on individuals and neo-Marxist approaches stressing structural features related to segmented labour markets (Preisendo¨rfer and Voss 1988). However, with the increasing interest in the concept of social capital, the structural features underlying social networks have been accentuated. Thus, the consensus among the main proponents of the concept is that social capital is a resource resulting from social structure (Bourdieu and Wacquant 1992: 119; Burt 2002; Coleman 1990). Hence, it /

2. Indeed, in some countries the pooling of family resources works both ways. In Spanish households with the head of the household out of work, employed youth residing with parents are found often to prevent overall household poverty by sharing their income with the rest of the family (Canto´-Sańchez and Mercader-Prats 2001).

557

EUROPEAN SOCIETIES

is not the personal relations between individuals or groups in themselves, but how these relations engage with the overarching institutional context that matters: ‘The same relations will be more or less productive under differing institutional constraints’ (Flap 2002: 41). The institutional context for the allocation of jobs may be structured by the nature and extent of state intervention in the labour market. It has been suggested that strong and comprehensive institutions of public employment services and active labour market programmes may partly offset the importance of personal social networks among job seekers. Thus, such institutions may improve the dissemination of information about job openings (Barbieri et al. 2000; Korpi 2001). We may also consider the impact of public institutions and policies on social relations operating in the job allocation process with reference to the theories of Bourdieu (1986). Following his approach, the valuation of different forms of capital depends on the particular institutional setting in which the form of capital is operated. In order to understand the structure and dynamics of differentiated societies, he argues, one needs to consider how the shape of specific ‘social spaces’, makes certain types of social networks valuable for the attainment of profits and privileges (Bourdieu and Wacquant 1992: 119). For Bourdieu, social networks can be more or less institutionalised. According to him, creating exclusive ‘occasions’ is a strategy pursued by elites to institutionalise social networks with the purpose to ensure that oneself and one’s offspring get to socialise with the right people in order to make valuable alliances (marriages, contractual arrangements, etc.), hence reproducing advantaged positions in the social structure (Bourdieu 1986). To develop this perspective a little further, welfare state interventions in the labour market may also be regarded as efforts to institutionalise social networks, but serving public rather than special interests. In fact, welfare states investing in active labour market policies may create important ‘occasions’, hereby democratising the access to social capital. Hence, by offering activation schemes (that is governmental schemes consisting of training, guidance, and/or subsidized employment) to the unemployed, occasions may be created where unemployed persons are brought together with employers and managers at employment or training sites. For the unemployed person undertaking the scheme, such meetings may lead to more permanent employment. Furthermore, a redistributive welfare state may counter the importance of social networks indirectly. Hence, it has been suggested that the degree of financial security among the unemployed may influence their job search behaviour, those with a low degree of financial security being more urged to utilise their ‘strong ties’ (Korpi 2001: 164). This phenomenon may be reinforced because of the interest of the ‘strong ties’ in helping the young unemployed person to become economically independent, at least to the 558


HARSLØF

extent that the economic burden related to the young persons subsistence rests on them. On the other hand, in societies where young unemployed people are generously supported by unemployment insurance, social assistance and other types of transfer incomes, this perspective suggests that social network relations have less importance in the job allocation process. Taken together, the theoretical considerations outlined above pertain to the way welfare states may affect recruitment processes through supply side interventions. However, demand side interventions may also affect the job matching process. In recent years the role of labour market protection legislation and how such legislation affects the composition of employment has received a great deal of attention (Esping-Andersen 2001). Relating this to recruitment processes, it can be argued that strict employment protection legislation may encourage employers to use informal recruitment channels as part of a comprehensive pre-screening procedure, due to the difficulties of terminating the employment of a poorly performing employee later on. Thus, in a strictly regulated labour market, one should expect a high extent of informal recruitment to prevail. Furthermore, informal recruitment patterns have been related to the general level of unemployment. It is argued that high unemployment prompts job seekers to exploit intensively all potential job contacts from their social network (Preisendo¨rfer and Voss 1988: 108). The level of unemployment is to a large extent dependent on international economic fluctuations beyond the control of governments. Still, some authors argue that the level of unemployment at least partly should be regarded as reflecting the political commitment to ensure full employment, which is argued to be an essential element of welfare state politics (Korpi and Palme 2000, 2003). The extent and design of such institutional arrangements employment services, activation schemes, welfare provisions for young unemployed, and employment protection legislation vary across Europe, as does the level of youth unemployment. It can therefore be expected that the importance of young people’s social network in the job allocation process varies widely between different European welfare states. The differences between the countries forming part of this study will be considered in section 4. First, the methodology of the study is presented. /

/

3. Data and measurements

This study is based on survey data from the European comparative Youth Unemployment and Social Exclusion study (YUSE see Hammer 2003). In the YUSE study, representative samples were drawn from national /

559

EUROPEAN SOCIETIES

unemployment registers and comprised young persons from the ages of 18 to 24 having experienced at least three months of continuous unemployment within a six-month observation period. In Germany and Spain, young persons between 25 and 27 years old were also included. Respondents were interviewed six to 12 months after the sampling had taken place. Surveys were fielded in the Nordic countries in 1995 1996, in Scotland in 1997 1998 and in Italy, Spain and Germany in 1999 2000.3 The data include the following national net samples (response rates in parentheses): Finland 1,736 persons (73 percent), Norway 1,106 persons (56 percent), Sweden 2,534 persons (63 percent), Denmark 1,171 persons (79 percent),4 Scotland 817 persons (56 percent), Spain 2,523 persons (52 percent), Germany 1,918 persons (60 percent) and Italy 1,421 (in Italy a cluster sample were carried out in two regions, Veneto and Campania for which response rates cannot be calculated). When operationalising the phenomenon of ‘informal recruitment’ in consideration of the problem under study, it was decided to apply a narrow definition. Only recruitment which was reported to have been facilitated by means of friends and/or relatives, i.e., the ‘strong ties’ in Granovetter’s conceptual framework (Granovetter 1995), was considered. This means that recruitment which has taken place because the employer has directly contacted the respondent, whom the employer may have known due to previous contacts, is not coded as informal recruitment (as long as the two are not considered by the respondent to be friends or relatives). This narrow definition applied here differs from Raffe’s (1985), who refers to recruitment, where a job-training scheme which is taking place at a firm is transformed into regular employment by the same employer, as a form of informal recruitment. The point is that by restricting the notion of informal recruitment to cases where friends or family are involved, one is allowed to study how active labour market policies may lessen the importance of having the right friends or family. Respondents who had at times had paid work were asked: ‘How did you find your current or most recent job?’ (henceforth referred to as the ‘reference job’). Respondents were presented with seven different answer /

/

/

3. The time difference between the different national surveys means that the surveys were fielded in different countries under different circumstances in terms of international economic fluctuations. However, as the macro-institutional variables, being most sensitive to such international trends / the youth unemployment rate and the youth activation rate / pertain to the specific years of the national surveys, this issue is deemed not to be a matter of concern for the present study. 4. The Danish sample was exclusively drawn from young people who were members of an unemployment insurance fund. This bias is accounted for in the multivariate analyses by the control variables on education, occupation and on whether the respondent has received unemployment insurance.

560


HARSLØF

categories: Jobcentre/labour market authorities, replied to advertisement/ placing advertisement yourself, contacted local employer, through union, through friends, through relatives, and other means (open category). Those reporting that they had obtained their reference job through either friends and/or relatives were coded ‘1’ in a dichotomous outcome variable. The question allowed multiple responses, and a substantial group reported combining different strategies to obtain their reference job. In cases involving a combination of formal and informal channels, respondents were categorised as having used an informal channel. Logistic regression techniques were applied to determine variance among welfare states in the prevalence of informal recruitment patterns in youth labour markets. Two groups of respondents were excluded from the study: those who reported that they had never been in paid employment, and those reporting that their reference job was being self-employed. A number of variables, which have been shown to affect the type of recruitment that people apply (Fevre 1989; Granovetter 1995), were included in order to account for individual differences. These control variables were gender, age, ethnic background (majority or minority),5 education, father’s education,6 total months as unemployed, and having children. Furthermore, the following variables relating to the reference job were included: sector (public or private), type of occupation (using the ISCO-88 classification system on a one digit level), work time (number of hours), and type of contract (temporary or permanent).7 A variable on whether or not the respondent had ever participated in a labour union meeting was included as a proxy of union membership.8 A variable on whether or not the respondent had received unemployment insurance in a 5. Having ‘Minority background’ was defined as being born in another country or having parents both of whom were born in another country. In Germany and Spain only respondents being themselves born in another country are categorised as belonging to an ethnic minority as there are no information on parents’ background. 6. The variables for education, including father’s education have three substantial categories: ‘Primary education/lower secondary’ /CASMIN code 1a/1b/1c/ 2a; ‘ Upper secondary education’/2b/2c; ‘Tertiary education’ /3a/3b. A fourth category was included to accommodate those answering ‘other type of education’ and the group of ‘missings’. 7. For those who were unemployed at the time of the survey, no information is available on the type of contract they had while occupying the reference job. This group is assigned to a third category encompassing those unemployed at the time of the interview. 8. Those respondents reporting that they have ever participated in a union meeting probably only constitute a subsection of those organised in unions. Still, in order to account for variance related to the reference job, it was considered relevant to adjust for this variable.

561

EUROPEAN SOCIETIES

12-month period prior to the interview was included as a proxy for membership of an unemployment insurance fund. Following Russell and O’Connell (2001), to further consider the welfare and labour market theories, additional regression models in which each country was replaced with its value on the different institutional variables, were set up. This strategy is in line with Przeworski’s and Teune’s (1970) methodological suggestion to treat social systems as residuals of theoretical variables in the search for explanations of social phenomena. However, with only eight countries, such exercises can only assist in exploring the field.9 The explanatory macro-institutional variables are presented in the following section.

4. Institutional differences among European countries structuring recruitment patterns

Table 1 presents institutional characteristics of the countries forming part of the study.10 Responding to the perspectives discussed in section 2, four main institutional factors are considered in the table. The first, active labour market policies (ALMP), is divided into two sub-sections, Public Employment Service (PES) and activation. Regarding PES, figures are presented on the share of GDP spent on PES divided by the national unemployment rate (column A). In this comparison Norway comes out as the largest spender on PES in proportion to the level of unemployment among the countries studied, followed by Sweden, Scotland, Germany, Denmark, Finland, Spain and Italy, the latter two countries appearing to spend very little. Of special importance to this study is the level of youth activation. By youth activation is referred to any kind of offers in terms of training, guidance and/or employment schemes provided for young unemployed people. An impression of the level of youth activation in the different countries can be obtained by regarding survey-based reports from young longer-term unemployed on whether they have participated in a 9. A test of co-linearity revealed one Tolerance value of 0.176 when all macro variables were included in the same model, a level which according to Menard (2002) should be a cause for concern. Therefore, no inferences are based on models including all macro variables in a full model. 10. In the case of Scotland, some of the figures presented cover the rest of the UK as well, as they were not exclusive to Scotland. However, according to Mooney and Poole (2004), social policy in Scotland up to the late 1990s has been very much in line with policies in England and Wales, following a trend of increasing social policy convergence during the Thatcherite period. Cash benefits are being provided on a uniform basis in Britain (English 1998).

562

Active labour market policies

Finland Norway Sweden Denmark Scotland/UK Italy Spain Germany

Unemployment compensation for youth

Public Employment Service (PES)

Activation

Expenditures as % of GDP divided by standardised unemployment rate. 1997 A

Percentage of young unemployed reporting participation in activation scheme at time of national survey B1

Expenditures on active labour market policies as% of GDP divided by unemployment rate. 1997 B2

0.01 0.037 0.029 0.02 0.0231 0.003 0.004 0.022

52 53 67 39 51 17 38 30

0.11 0.24 0.23 0.27 0.051 0.08 0.03 0.12

Average of the standardised figures in column A, B1 and B2

/0.164 1.074 1.115 0.411 /0.041 /1.214 /0.917 /0.265

Proportions of young unemployed reporting receipt of transfer income within 12-month reference period

Median of reported net monthly income among young unemployed by time of national survey (Purchase Power Standards)

C1

C2

92 87 89 96 84 5 21 78

322 401 502 684 225 229 390 453

Average of the standardised figures in columns C1 and C2

0.066 0.256 0.619 1.320 /0.369 /1.476 /0.717 0.301

Strictness of regular employment protection legislation. Late 1990s

Youth unemployment (15 24 years old) at time of national survey

D

E

2.3 2.3 2.9 1.5 0.91 1.8 2.6 2.7

25.3 10.6 21.0 8.1 21.31 31.5 25.5 7.7

563

HARSLØF

A. Own calculations. Sources: OECD Social Expenditure Database (OECD 1998). The standardised unemployment rate applied in the calculation for Italy refers to 1996. B1. Generated on the basis of the YUSE data set (for the procedure applied see Harsløf 2005). B2. Own calculations. The unemployment rate used in the calculation for Italy refers to 1996. Sources: OECD Social Expenditure Database (OECD 1998). C1. Generated on the basis of the YUSE data set. Unemployed are defined as those reporting to be unemployed or enrolled in a government activation scheme in a reference week prior to the time of the survey. Transfer income includes: unemployment insurance, social assistance (flat rate), sickness benefit, student grant, housing benefit, and payments related to scheme participation. C2. Generated on the basis of the YUSE data set. Unemployed are defined as those reporting to be unemployed or enrolled in a government activation scheme in a reference week prior to the time of the survey. The median of self-reported disposable amounts are converted to Purchase Power Standards. D. Source: OECD 2004: 117. E. Source: OECD 2002: 24. 1 Figures for the United Kingdom.


TABLE 1. Macro institutional variables: Public Employment Service, activation, unemployment compensation, strictness of regular employment protection legislation and level of youth unemployment

EUROPEAN SOCIETIES

governmental scheme for unemployed (column B1). The Nordic countries, with Sweden uppermost, have the highest shares of young unemployed persons reporting participation in activation. In this comparison, Germany lies somewhat below the average, whereas Scotland appears to maintain a level of activation on a par with the Nordic countries. Regarding the Southern European countries, we find that a rather large proportion of Spanish unemployed youth report having participated in an activation scheme, whereas Italy stands out as that country with the fewest young activation scheme participants. As a second indicator, the share of ALMP expenditures on GDP is divided by the national unemployment rate. This indicator relates to all age groups and not the group of young unemployed in particular.11 Using this indicator, Denmark, Norway and Sweden are uppermost; Germany and Finland form a middle group; Italy, Spain and Scotland as part of UK spend the least on ALMP in proportion to their unemployment rate. In order to summarise these ALMP indicators in a single measure, the variables were standardised. The average of these three standardised indicators is presented in the table. Sweden is on top, being on average 1.115 standard errors above the mean, whereas Italy, being 1.214 standard errors below the mean, is at the bottom in this summarising indicator. Regarding unemployment compensation for young unemployed, two indicators are considered. Firstly, the proportion of young unemployed reporting that they have received any type of public transfer income during a 12-month reference period are presented (column C1). By using such an inclusive question, this indicator is rather rough. Nevertheless it provides results consistent with findings from other comparative studies (cf. Russell and O’Connell 2001). Secondly, median net monthly income among the groups in the Youth Unemployment and Social Exclusion data, who were unemployed at the time of the survey, was calculated for each country. The amounts, converted into Purchase Power Standards (PPS) in order to correct for a harmonised consumer price index, are presented in column C2. According to these figures, unemployed young people in Denmark and Sweden enjoy the highest disposable income each month. However, Spanish and German unemployed respondents reported relatively high incomes on a par with the Nordic countries with the exception of Denmark. This may be due to the fact that self-reported net income may include income from informal sources, such as work in the informal economy or contributions from parents. The Italian and Scottish 11. The OECD public expenditure data base includes an entry on ‘Youth measures’. This, however, is inadequate for our purpose as it only covers special programmes, and not young people’s participation in programmes open to all age groups (Martin and Grubb 2001: 13).

564


HARSLØF

unemployed respondents were worst off according to this indicator. The figures in these two indicators have been standardised and the average is presented in the table as a summarising indicator on the economic protection of young unemployed. The OECD measure of the strictness of employment protection legislation covers three aspects: the legislation concerning regular (individual) employment, collective dismissals, and temporary employment. In the table (column D), only OECD’s rating of the former aspect is considered. This aspect is regarded to be the one which is of relevance when evaluating the theory of high dismissal costs prompting employers to use informal contacts in the recruitment procedure to screen candidates. This aspect is again made up of three indicators: the regular procedural inconveniences related to dismissal, the notice and severance pay for nofault individual dismissals, and the difficulty of dismissal. When these are combined, Sweden is rated as the country with the strictest regular employment protection legislation, followed by Germany and Spain. Finland and Norway have a somewhat less strict legislation. Italy and Denmark are in-between, while the UK stands out as having relatively lax legislation. The level of youth unemployment varies among the countries considered (column E). The figures for the year of the national surveys are used so that figures for the Nordic countries plus Scotland cover 1997 and the figures for the remaining countries cover 2000. The highest youth unemployment rate is found in Italy, followed by Spain and Finland, again followed by Scotland/UK and Sweden. Norway, Germany and Denmark are the countries with the lowest youth unemployment rate at the time of the national survey. According to the theoretical perspectives outlined in section two pertaining to welfare state supply side intervention, it is to be expected that informal recruitment channels are of greatest importance in countries with low spending on employment services, low level of activation and low degree of economic protection of young unemployed. This perspective suggests high informal recruitment in countries such as Italy and Spain, and low levels in the Nordic countries. A country such as Scotland, offering a medium level benefit coverage while providing a relatively high youth activation level could, according to the theories, be expected to have an intermediate level of informal recruitment when compared to the other countries. However, regarding the perspective stressing welfare state demand side intervention, informal recruitment should be widespread in the Swedish youth labour market with its very strict employment protection legislation, while less widespread in liberal labour markets, such as the Scottish and the Danish. As for youth unemployment, the high youth 565

EUROPEAN SOCIETIES

unemployment rates observed in countries including Italy, Spain, Finland and Sweden give reason to expect a high proportion of employment relationships resulting from processes of informal recruitment.

5. Informal channels into the labour market in different European welfare states

By the time of the survey, 37 percent of the young people in the YUSE study were in paid employment. Of the remaining group, however, a substantial group had previously been in paid employment, so that, in total, 78 percent of the respondents could refer to a situation where they had acquired a job. Figure 1 reports proportions of young people with a reference job, and the proportions of this group reporting that they got their reference job by means of friends and/or relatives. Of the eight countries, Denmark is the country in which informal recruitment is least frequently reported with Italy as the other extreme.12 A series of logistic regression analyses were carried out to investigate the theoretical perspectives. Four explanatory variables were constructed pertaining to these perspectives. Thus, to explore the different perspectives, one model considered active policies (PES and ALMP). The summarising measure of PES and activation (i.e., the average of the standardised figures in columns A, B1 and B2 presented in Table 1) was used here (Model 2). Similarly, the summary measure of the level of income protection enjoyed by young unemployed persons in the respective countries (i.e., the average of the standardised figures in column C1 and C2 presented in Table 1) was used in Model 3. Together, these two models accommodated the theoretical perspective that the prevalence of informal recruitment may be affected by supply side policies. The demand side perspective was considered with two additional models. Firstly, in Model 4, country dummies were replaced with the OECD ranking of the strictness of employment regulation (cf. column D, Table 1). Secondly, Model 5 used the level of youth unemployment (cf. column E, Table 1). To produce comparable estimates, the two variables covering the demand side perspective were also standardised. Table 2 presents the results in terms of odds ratios. Starting with the control variables common to all models, as can be seen, the informal 12. 83 percent of the respondents from Campania (a region in the Southern part of Italy) against 77 percent of the respondents in Veneto (in the Northern part), report being recruited by informal means, a difference significant on a 5-percent level. Regarding Germany, respondents from the Eastern counties (61 percent) report having used informal channels to a significantly higher extent than their Western counterparts (55 percent).

566


HARSLØF

95

Denmark

30 80

Sweden

37 62

Finland

39 77

Scotland

39 79

Norway

40 77

Spain

50 78

Average (unweighted)

51 76

Germany

57 77

Italy

81

0

20

Informal recruitment

40

60

80

100

Ever been in paid employment

Figure 1. Proportions of respondents ever been in paid employment and the proportions of this group reporting that they obtained their reference job through friends and/or relatives. N /13,226. Source: YUSE data.

recruitment channels are especially important when it comes to jobs for the youngest job seekers. Furthermore, they are most important for young males for females, the odds of being informally recruited deceases with 20 percent. Such gender difference has also been reported in other studies (Lin 2001), and may be explained by the existence of gender homophilia in social networks (Marsden 1987). Across Europe, female participation in the labour market is lower than male participation (OECD 2003: 83). If the social networks of young women primarily comprise other women the total labour market attachment of such networks will on average be weaker than the social networks of young men. In three of the models, having children significantly increases the odds of getting the job by informal means. This may be due to the fact that having children bring the young unemployed closer to other relatives, who may put their social networks at the disposal of the young unemployed.13 Accumulated unemployment appears not to be associated with the outcome variable. Having received unemployment insurance benefit within a 12-month reference period is negatively associated with having used informal recruitment channels. This may reflect differences in the labour markets for insured and noninsured labour, the former being more formalised. The analyses further suggest that informal recruitment channels are applied primarily in hiring /

13. American research has found that childrearing affects the social relations of men and women in quite different ways. The male’s social network tends to be enhanced in the sense that they are drawn into greater contact with other family members. The network of females, on the other hand, tends to be disrupted / their network size and contact volume has been found to be reduced (Munch et al . 1997).

567

EUROPEAN SOCIETIES TABLE 2.

Logistic regression analyses of being recruited through informal channel. Odds ratios Model 1

Model 2

Model 3

Model 4

Model 5

Woman Age (cont.) Ethnic minority Having children

0.804*** 0.942*** 1.049 1.179*

0.779*** 0.916*** 1.102 1.180*

0.785*** 0.927*** 1.142 1.221**

0.786*** 0.939*** 1.093 1.076

0.778*** 0.950*** 1.189 1.151

Primary/lower secondary education Higher secondary education Tertiary education Other education or missing

Ref.

Ref.

Ref.

Ref.

Ref.

0.919 0.792** 0.915

0.926 0.939 0.893* 0.875* 0.706*** 0.715*** 0.698*** 0.676*** 0.903 0.856 0.865 0.806*

Ref.

Ref.

Ref.

Ref.

0.953 0.794** 1.006

0.957 0.807** 0.994

1.042 0.877** 1.081

0.902 0.938 0.755*** 0.785** 0.910 0.963

1.000

1.002

0.999

1.001

0.931

0.738*** 0.997

0.655*** 0.826***

Permanent employment Ref. Fixed term employment 1.093 Not employed by time of survey 0.992

Ref. 1.105 1.003

Ref. 1.051 1.010

Ref. 1.150* 1.016

Ref. 1.101 1.030

Working 20 h or less 2130 h 3140 h 4150 h More than 50 h or missing

Ref. 0.754** 0.685*** 0.828 0.717**

Ref. 0.718*** 0.574*** 0.768** 0.658***

Ref. 0.767** 0.675*** 0.812* 0.641***

Ref. 0.674*** 0.558*** 0.785* 0.609***

Ref. 0.665*** 0.565*** 0.773** 0.621***

Public sector

0.653*** 0.558*** 0.573*** 0.595*** 0.578***

Elementary job Managers Professionals Technicians, associate professionals Clerks Service, shop/market sales workers Skilled agricultural/fishery workers Craft and related workers Plant operators and assemblers Other or missing

Ref. 0.465** 0.441*** 0.509***

0.708*** 0.795** 0.702*** 0.802* 0.971 1.339*

0.746*** 0.789** 0.751** 0.745** 1.176 0.954

0.768** 0.724*** 1.002

Participated in union meeting

0.859*

0.943

0.870

Father’s edu.: Primary/lower secondary Father’s edu.: Higher secondary Father’s edu.: Tertiary Father’s edu.: Other/missing Total months unemployed (cont.) Unemployment insurance

568

Ref. 0.508* 0.576*** 0.576***

Ref. 0.552* 0.487*** 0.540***

Ref. 0.504* 0.576*** 0.590***

Ref.

0.999

Ref. 0.552* 0.515*** 0.574***

0.558*** 0.723*** 0.624*** 0.717*** 0.671*** 0.687*** 0.841** 0.771** 0.834* 0.802** 0.892

1.051

0.897*

0.986

0.973

0.848*

0.958


HARSLØF

TABLE 2 (Continued)

Model 1 Finland Norway Sweden Denmark Scotland Spain Italy Germany PES/Activation-measure, stand. (cont.) Income protection-measure, stand. (cont.) Strictness of EPL-indicator, stand. (cont.) Youth unemployment, stand. (cont.) Constant N Model x2a Df

Model 2

Model 3

Model 4

Model 5

Ref. 1.502*** 2.088*** 0.805* 2.443*** 1.866*** 6.058*** 1.739* 0.710*** 0.533*** 1.009 1.325*** 6.508*** 7,254 815*** 36

18.59*** 7,254 476*** 30

13.01*** 7,254 648*** 30

12.6*** 7,254 440*** 30

9.645*** 7,254 525*** 30

***P B/0.001; **P B/0.01; *P B/0.05. aInitial model without country dummies: x2 440***, df: 29. Baseline: Male, 18 years old, belonging to ethnic majority, no children, primary/lower secondary education, father: primary/lower secondary education, three months total unemployment, not received unemployment insurance within the last 12 months, permanent employment, working 20 hours or less, private sector, elementary job, Finland (Model 1), PES/Activation / /1.214 (Model 2), Unemployment protection / /1,476 (Model 3), Strictness of EPL /0.9 (Model 4), youth unemployment /7.7 (Model 5).

for jobs with few qualification requirements. Thus, respondents with only primary or lower secondary education tend to be more likely to report having entered the labour market with the help of friends and/or relatives, even if this relationship, apart from Models 4 and 5, is only statistically significant when this group is compared to the group with tertiary education. Information on the educational level of the respondent’s father can also assist in predicting the outcome variable. Thus, a respondent whose father has higher education tends to be less likely to report having used an informal network when getting the job. This indicates that the phenomenon of informal recruitment is related to lower classes in the social strata. This perspective is corroborated when regarding information about the reference job. Having elementary jobs, the lowest ranking main category in the ISCO 88 occupational hierarchy, as the reference, all higher ranking job categories are related to reduced probability for being recruited informally. According to Granovetter, this pattern is probably 569

EUROPEAN SOCIETIES

partly explained by the fact that employers looking for unskilled labour have a greater interest in having reliable information on the job seekers’ ‘soft skills’. Information on such skills may be obtained by recruiting among personal network or among the networks of those already employed at the firm (Granovetter 1995: 171). Apart from Model 4, the control variable on type of contract is not found to be significant. It is primarily the private sector which recruits by way of networks. This probably has to do with the fact that the private sector is not subject to the same strict legislation regarding recruitment procedures as is the public sector (Hagtvet 2005). Having participated in a union meeting is negatively related to informal recruitment. This could indicate that workplaces with organised labour tend to counteract informal recruitment processes, a pattern also found in other studies (cf. Fevre 1989: 93). Having controlled for these variables, some individual country effects emerge. Introducing country dummies to the model significantly improves the fit of the model, indicating that recruitment patterns not only relate to individual or job related factors but to the dynamics of larger social systems as well. Compared to Finland (the omitted reference category), Italy stands out as the country in which young people are most dependent on their social network in the job search process. Young Italian employees from the sample had an increased probability, by a factor of 6, of having entered this job position by means of informal networks when compared to young Finnish persons. Model 1 points to substantial differences among the Nordic countries. Danish respondents report less frequent recruitment through informal channels compared to their Finnish counterparts. The Swedish respondents, on the other hand, are more likely to be employed in this manner. Model 2 replaces country dummies with the summary ALMP-measure. It shows that for each unit increase in the explanatory variable in terms of standard deviations, the odds of being employed informally are reduced by 29 percent. In Model 3, a variable representing the level of social protection of young unemployed persons in the various countries replaces country dummies. This variable also has the direction expected. By using the strictness-of-EPL variable applied in Model 4, on the other hand, no substantial effect is identified. Using the level of youth unemployment at the time of the national survey as a substitute for country dummies, as done in Model 5, a positive relationship appears: the higher the level of youth unemployment in the country where the respondent resides, the higher the odds of entering the labour market through informal channels. By considering the fit of the models one can further explore the explanatory power of the theoretical perspectives on the national incidence of informal recruitment. By comparing the fit of the mo570


HARSLØF

del without country dummies or other macro institutional variables (chi-square 440) with Models 1 and 2, it may be calculated that about 10 percent of the total cross-national variation can be attributed to the variable on active labour market policies ((476 440)/(815 440) 0.096). By the same procedure, 55 percent of the total cross-national variation is shown to be related to the variable on social protection ((648 440)/ (815 440) 0.554), and about 20 percent to the level of youth unemployment ((525 440)/(815 440) 0.226). As the effects of the different explanatory variables on the dependent variable may partly overlap with one another, one has to be careful in drawing conclusions. Nonetheless, the results are consistent with the theoretical argument that welfare state interventions, especially income replacement provided for young unemployed, have an impact on the degree of formalization of the job matching process in youth labour markets. /

/

/

/

/

/

/

/

/

/

6. Conclusion

The analysis in this article indicates a certain profile of the types of jobs which are occupied through informal recruitment. Jobs occupied in this manner are more frequently in the private than the public sector. They are more often part-time than full-time. Informal recruitment is particularly prevalent in the market for unskilled labour. Thus, to a certain degree, the analyses disclosed a relationship between using informal channels and having a rather marginal position within the labour market. Using one’s social ties to obtain a paid job appears as a residual strategy for the most marginalised groups of young unemployed. If this runs counter to the notion of social capital as a resource for improving one’s life chances, one has to take into consideration that the counterfactual event could be getting no job at all. In this perspective, even a part-time job at the lower end of the occupational hierarchy, acquired by means of friends or relatives, may constitute an important improvement of the young person’s social situation. Across the European countries considered in this article, the web of social ties unemployed young people command seem to be important for their chances of getting a foothold in the labour market. This insight supports conventional wisdom that ‘it’s who you know’ that matters when it comes to finding a job. However, the analyses have suggested that the way social networks operate is not a given, universal fact. This is seen by reference to the substantial country variance in the prevalence of informal recruitment found even when a range of individual and job related variables are accounted for. Furthermore, in this article the phenomenon of informal recruitment has been related to structural features emerging 571

EUROPEAN SOCIETIES

from the institutional context provided by welfare state intervention in the labour market. Thus, the importance of the informal channel of recruitment appears to be related to the extent of active labour market policies, the economic provision of benefits and social assistance for young unemployed, and the level of youth unemployment. Indeed, the job matching process, as studied in these eight European countries, appears as a structured process; individual strategies to escape unemployment in the phase of youth is affected by structural features originating in the relationship between state, market and civil society. It appears that generous, redistributive welfare states not only decommodify human labour (Esping-Andersen 1990), but to a certain extent also personal relations between people. With only eight countries in the study such conclusions can only be tentative. Considering the important problem of marginalised youth across Europe, more comparative research addressing the institutional effects on recruitment processes is needed. Questions on the channels of recruitment which job seekers apply ought to be included in larger European-wide surveys, to allow for comprehensive comparative analyses. That processes of recruitment in youth labour markets can be influenced by welfare state policies, could have implications for European strategies to counter the problem of youth unemployment. For the European Union combating social exclusion has a high priority, and special attention is paid to people’s connection with the labour market (Hunt 2005). In this respect, one may emphasise the suggestion made by Windolf (1986), that in determining processes of social exclusion, the subtle processes of recruiting labour should be considered as even more important than the more overt processes of dismissing labour. Until now, efforts for promoting ‘equal access’ in the job allocation process, has been more focused on the recommendation of positive (affirmative) action measures for certain disadvantaged groups than addressing how prevailing recruitment practices may produce labour market inequalities in member countries (cf. Wahl 2005). Here, it may be fruitful to consider how to ‘democratise’ social capital for young people who are about to enter the labour market. Whereas encouraging and, to some extent, even financing youth activation programmes is already on the European Union agenda, co-ordinating strategies to strengthen the welfare provisions for young unemployed persons appears to be a strategy worth considering to achieve more equal opportunities among those on the margin of the labour market.

572


HARSLØF

Acknowledgements

The author wishes to thank Torild Hammer, Axel West Pedersen, Kirsti Valset (Norwegian Social Research), Einar Øverbye (Oslo University College), Lars Mjøset (University of Oslo), and three anonymous reviewers for comments on earlier drafts of this paper.

References

Barbieri, P., Paugam, S. and Russel, H. (2000) ‘Social capital and exits from unemployment’, in D. Gallie and S. Paugam (eds), Welfare Regimes and the Experience of unemployment in Europe, Oxford/New York: Oxford University Press. Bison, I. and Esping-Andersen, G. (2000) ‘Unemployment, welfare regime, and income packaging’, in D. Gallie and S. Paugam (eds), Welfare Regimes and the Experience of unemployment in Europe, Oxford/ New York: Oxford University Press. Bourdieu, P. (1986) ‘The forms of capital’, in J. G. Richardson (ed.), Handbook of Theory and Research for the Sociology of Education, New York: Greenwood Press. Bourdieu, P. and Wacquant, L. J. D. (1992) An Invitation to Reflexive Sociology, Cambridge: Polity Press. Burt, R. S. (2001) ‘Structural holes versus network closure as social capital’, in N. Lin, K. Cook and R. S. Burt (eds), Social Capital: Theory and Research, New York: Aldine de Gruyter. Burt, R. S. (2002) ‘The social capital of structural holes’, in M. F. Guilleń, R. Collins, P. England and M. Meyer (eds), New Directions in Economic Sociology, New York: Russell Sage Foundation. Canto´-Sańchez, O. and Mercader-Prats, M. (2001) ‘Young people leaving home: the impact on the poverty of children and others in Spain’, in B. Bradburry, S. Jenkins and J. Micklewright (eds), The Dynamics of Child Poverty in Industrialised Countries, Cambridge: Cambridge University Press. Coleman, J. (1990) Foundations of Social Theory, Cambridge, MA: Harvard University Press. English, J. (1998) ‘Social security’, in J. English (ed.), Social Services in Scotland, 4th edition, Trowbridge: The Cromwell Press. Esping-Andersen, G. (1990) The Three Worlds of Welfare Capitalism, Cambridge: Polity Press.

573

EUROPEAN SOCIETIES

Esping-Andersen, G. (2001) ‘Who is harmed by labour market regulaitons? Quantitative evidence’, in G. Esping-Andersen and M. Regini (eds), Why Deregulate Labour Markets, Oxford: Oxford University Press. Fevre, R. (1989) ‘Informal practices, flexible firms and private labour markets, Sociology 23 (1): 91 109. Flap, H. (2002) ‘No man is an island’, in O. Favreau and E. Lazega (eds), Conventions and Structures in Economic Organization. Markets, Networks and Hierarchies, London: Edward Elgar. Granovetter, M. (1985) ‘Economic action and social structure: the problem of embeddedness’, American Journal of Sociology 81 (3): 489 515. Granovetter, M. (1988) ‘The sociological and economic approaches to labor market analysis: a social structural view’, in G. Farkas and P. England (eds), Industries, Firms, and Jobs: Sociological and Economic Approaches, New York: Plenum Press. Granovetter, M. (1995) Getting a Job: A Study of Contacts and Careers. Afterword to the 2nd edition, 2nd edition, Chicago, IL: University of Chicago Press. Hagtvet, H. (2005) Undersøkelse om rekruttering av arbeidskraft [Study of the recruitment of labour power], No. 1. Oslo: Aetat. Hammer, T. (2003) Youth Unemployment and Social Exclusion in Europe, Bristol: Policy Press. Harsløf, I. (2005) ‘‘‘Integrative’’ or ‘‘defensive’’ youth activation in nine European welfare states’, Journal of Youth Studies 8 (4): 461 81. Hunt, J. (2005) ‘Combating social exclusion: The EU’s contribution, Journal of Social Welfare and Family Law 27 (1): 113 20. Korpi, T. (2001) ‘Good friends in bad times? Social networks and job search among the unemployed in Sweden, ACTA Sociologica 44: 157 70. Korpi, W. and Palme, J. (2000) ‘Distributive conflict, political mobilization and the welfare state: Comparative patterns of emergence and retrenchment in the Westernized countries. Paper presented at the annual meeting of the American Sociological Association, Washington DC. Korpi, W. and Palme, J. (2003) ‘New politics and class politics in the context of austerity and globalization: Welfare state regress in 18 countries, 1975 95’, American Political Science Review 97 (3): 425 46. Lin, N. (2000) Inequality in Social Capital, Durham: Duke University. Lin, N. (2001) Social Capital: A Theory of Social Strucutre and Action, Cambridge: Cambridge University Press. /

/

/

/

/

/

574

/


HARSLØF

Marsden, P. (1987) ‘Core discussion networks of Americans’, American Sociological Review 52: 122 31. Martin, J. P. and Grubb, D. (2001) ‘What works for whom: A review of OECD countries’ experience with active labour market policies’, Swedish Economic Policy Review 8: 9 56. Menard, S. (2002) Applied Logistic Regression Analysis, 2nd edition, London: Sage. Mooney, G. and Poole, L. (2004) ‘‘‘A land of milk and honey’’? Social policy in Scotland after Devolution’, Critical Social Policy 24 (4): 458 83. Munch, A., Smith-Lovin, L. and McPherson, J. M. (1997) ‘Gender, children, and social contact: The effects of childrearing for men and women, American Sociological Review 62: 509 20. OECD (1998) Employment Outlook, Paris. OECD (2002) Economic Outlook, Paris. OECD (2003) Employment Outlook: Towards More and Better Jobs, Paris. OECD (2004) Employment Outlook, Paris. Portes, A. (1998) ‘Social capital: its origins and applications in modern sociology’, Annual Review of Sociology 24 (1): 1 24. Preisendo¨rfer, P. and Voss, T. (1988) ‘Arbeitsmarkt und soziale Netzwerke: Die Bedeutung sozialer Kontakte beim Zugang zu Arbeitspla¨tzen, Soziale Welt 39 (1): 104 19. Przeworski, A. and Teune, H. (1970) The Logic of Comparative Social Inquiry, New York: Wiley-Interscience. Raffe, D. (1985) ‘Degrees of informality: Methods of job placement among Scottish school-leavers, British Journal of Guidance and Counselling 13 (3): 292 307. Russell, H. and O’Connell, P. (2001) ‘Getting a job in Europe: The transition from underemployment to work among young people in nine European countries, Work, Employment & Society 15 (1): 1 24. Wahl, A. V. (2005) ‘Liberal, conservative, social democratic, or . . . European? The European Union as equal employment regime, Social Politics 12 (1): 67 95. Windolf, P. (1986) ‘Recruitment, selection, and internal labour markets in Britain and Germany’, Organization Studies 7 (3): 235 54. /

/

/

/

/

/

/

/

/

/

575

EUROPEAN SOCIETIES

Ivan Harsløf is associate professor and managing director of the Research Group for Inclusive Social Welfare Policies, Oslo University College. He has previously worked at the Danish National Institute of Social Research, and as doctoral student at Norwegian Social Research. During the spring of 2006 he was a Marie Curie RTN fellow at the Department of Sociology and Social Research, University of Milan-Bicocca. In his dissertation, YoungMan’s-Land: European Youth Between Welfare and Work, he compares social and labour market policies aiming at integrating young unemployed persons into the labour market across different European welfare states.

Address for correspondence: Ivan Harsløf, Oslo University College, Research Group for Inclusive Social Welfare Policies, P.B. 4, St. Olavs plass, NO-0130 Oslo, Norway. Tel: /47 22 45 36 88. Fax: /47 22 45 30 65. E-mail: [email protected]. Homepage: www.hio.no/giv

576