Informal Recruitment Methods and Disadvantages of Immigrants in the Swedish Labour Market Alireza Behtoui†

† Alireza Behtoui is a research fellow at the Department of Thematic Studies (Ethnic Studies), at Linköping University. He can be reached at: Linköping University, Department of Thematic Studies, Campus Norrköping/ITUF, SE-601 74 Norrköping, SWEDEN, and, The National Institute for Working Life, Laxholmstorget 3, 602 21 Norrköping. E-mail: [email protected] The author is grateful to Carl-Ulrik Scheirup, Mahmood Arai and Anders Neergaard for their helpful comments and suggestions on earlier drafts of this paper.

1

Abstract

This paper examines the probability of success of different methods for finding employment, both in times of economic growth and decline in the Swedish labour market. Most previous research about the importance of informal hiring practices in job acquisition is about the United States’ largely unregulated labour market, in which social networks would be expected to matter, while job seekers in Sweden have a substantively important alternative information channel in the public employment agencies. Data set on hiring processes, among West European studies, have not allowed researchers to examine the relationship between income and job contacts among various groups of immigrants. Due to the enough large data set, this article examines the job-search strategies and labour market outcomes among individuals in two large groups of immigrants and natives. Two questions are examined; do immigrants and natives differ in their job-search strategies? Does the effect of job-search strategies vary for each group? The findings show that natives rely more on informal search strategies than do immigrants. The use of informal search strategies leads to slightly higher-paying jobs for natives but to considerably lower-paying jobs for some immigrants.

Keywords: Job search methods, Social network, Immigrants

2

Introduction

The study of job search has gained much attention in recent years, with a special concern for the efficiency of different search methods in getting employment. This line of work emphasizes the role social networks play in labor market processes. It links the use of informal strategies for job search to labor market outcomes such as wage or social status of the position(Granovetters 1995; Marsden and Gorman 2001). Most previous research about the importance of informal hiring practices in job acquisition is about the United States’ largely unregulated labor market, in which social networks would be expected to matter. This article investigates the effect of social networks in Sweden, which has a substantively important alternative information channel in the public employment agencies.

Lack of access to valuable social networks to get a job is one plausible mechanism for explaining the labour market problems of disadvantaged minority groups. Studies of the role of social networks in the labour market for minority groups and immigrants have focused on attainment of jobs and have found support for the importance of social networks as an explanation for the inferior labour-market position of these groups (Ports 1993, Marx and Liecht 1992, Petersen et al. 2000).

3

This paper studies to what extent natives (Swedish-born individuals) and immigrants (foreign-born individuals) use their social networks to find a job and how social networks are related to wage differences between employed immigrants and natives in the Swedish labor market. Unfortunately, most data on job-search strategies and hiring processes, among West European studies, have been too small and not allowed researchers to compare various immigrant groups and natives regarding job-search strategies. Due to the enough large data set, this article examines the impact of social networks on the labor market outcomes among individuals in two large groups of immigrants and natives.

Immigrants in this study are defined as individuals born outside Sweden. Here we single out two different regions of origin; firstly, North-West countries, which include North-West Europe and North America (NW)1 and secondly Outside of North-West Europe and North America (ONW), which is rest of the world. This distinction was made because, according to earlier studies, individuals from the ONW countries have an inferior status compared to individual from the NW countries in Swedish labour market and run a higher risk of suffering from discrimination (Wadensjö 1997; Arai, Regner & Schröder 2000; Arai et al. 2000; leGrand & Szulkin 2001, and Nekby 2001).

4

More specific, the paper intends to examine the following issues; • Is there any difference in the use of informal methods between natives and immigrants? In the light of research on job-search methods, we expect that immigrants to be less likely to find a job through informal search methods. • We will examine returns to various methods of search. We expect that in general, methods involving use of social networks will be most likely to lead to better paid jobs. • According to previous studies (for example Green et al. 1999), disadvantaged minority groups may find job via their own group ties, but the lower the average income is in one’s immigrant group, the less advantage in income from using one’s same-group contacts, compared to other methods of finding job. Hence, the returns to informal job-search methods are expected to be poorer for ONW immigrants than for natives. • The economic conditions provide the context in which the search takes place, and therefore are important in determining the availability of jobs in general, and adequate jobs in particular. We investigate the impact of the economic conditions on choosing different search methods.

The empirical results of this paper indicate that it is less likely that immigrants in general have found their jobs through personal contacts, in comparison with natives. Our results furthermore suggest that the use of informal search strategies

5

leads to slightly higher-paying jobs for natives but to considerable lower-paying jobs for immigrants from ONW countries. Empirical results also demonstrate that, at the time of economic recession informal methods are used, more frequently than other methods, to find a job.

We begin in the next section by defining job-search methods, presenting a theoretical background regarding minorities and job-search methods, and a review of previous studies in this field. Thereafter we describe our data, the measures and the empirical set-up. The empirical results are reported as follows: first, the variation in job-search strategies among different groups, second, the effect of the macro economic conditions on job-search strategies and, third, the impact of informal job-finding methods on wages. The paper concludes with a summary and a discussion.

6

Job-search methods

In the job-search methods literature, three ways of finding a job are distinguished as the most common methods: 1. Formal methods of job-search including public and private employment agencies along with newspaper advertisements. The defining characteristic of formal means is that the job seeker uses the services of an impersonal intermediary, between himself and prospective employers. 2. Informal methods of job-search or social networks including help from friends and relatives or contact with previous co-workers and employers. According to Granovetter (1974:11), this way of finding a job “ implies that there is some individual known personally to the respondent, with whom he originally became acquainted in some context unrelated to a search for job information”. 3. Direct applications means that one makes direct contact with a firm, through visit or through writing, and that one does not use a formal or personal intermediary, and has not heard about a specific opening from a personal contact. Mardsen & Gorman (2001:469) point out that personal network may or may not be involved in the case of “direct application”. Granovetter (1974:11) notes that “direct application” has in some studies been included with personal contacts and some times has been seen as a formal method, but “in fact, it is different from both formal methods and from the use of personal contacts”.

7

Minorities and job-search methods

Social segregation or discrimination in contacts of disadvantaged minority groups may manifest itself in, inter alia, the “choice” of social intimates, neighbors, schools and friends, while discrimination in contracts means the unequal treatment of otherwise similar persons on the basis of their skin colour, hair colour, or name (Loury 2002). In many ways, the effects of discrimination in contracts and

discrimination in contacts (social segregation) are interwoven.

The fact that people to a great extent find jobs via information supplied through social networks may serve as a mechanism, which preserves particularism (Coverdill 1998). Some individuals have the right contacts, while others do not and if one lacks appropriate contacts, there is little one can do about it (Granovetter 1974:16). Social networks tend towards closure along most dimensions; individuals hired through networks tend to be socially similar to those referring them (Mardsen 1988). The flip side of a beneficial social integration, due to preexisting social ties within a specific community, is the cost of exclusion of outsiders. The use of informal methods to identify and select employees leaves an employer vulnerable to the charge of prejudice and favoritism (Marsden & Gorman 2001: 482). Braddock and Mcpartland (1987) name segregated recruitment networks as one of four primary exclusionary barriers, which works against achieving equality of employment opportunities for minorities. Manwaring (1984) and Fevre (1989)

8

report that use of network-based staffing procedures tends to reproduce patterns of particularism and social closure. Holzer, moreover contends that informal recruitment strategies may permit race to play a more significant role in the hiring process, which would thus hurt minorities’ chances of getting good jobs. His research demonstrates that “black” Americans are more likely to obtain jobs when they use formal methods, because formal methods provide more explicit criteria by which employers can judge candidates (Holzer, 1987, 1988). His findings are supported by Mier and Giloth’s study (1985) of labor market practices in a Hispanic area of Chicago and Marx and Leicht’s analysis (1992) of job-search strategies among manufacturing employees in Indiana

Differences between minority groups and “whites” in employment rate appear, in part, to be attributed to differences in contacts (Holzer 1987). Osterman (1980) reports that 57 percent of “whites” found their jobs through social networks while for blacks this was the case for only 33 percent. As Green et al. (1995) report the networks of the poor are less apt than those of nonpoor to include employed persons.

Positions in which minorities are concentrated typically offer lower levels of income and status than jobs dominated by majority. Therefore it is plausible to expect that use of contacts might yield lower wage returns to minorities because their social networks provide less access to high-paid jobs than do those of whites.

9

A number of U.S. studies examined variation by race in the wage effects of search method use. By virtue of segregating processes and homophily2 in networks, the social networks of black Americans are less apt to include people in a position to pass along private job information or to make hiring decisions. Holzer (1988) suggested that the networks of blacks might be less productive than those of whites by virtue of household composition, high male unemployment, and lower willingness of white employers to rely on referrals and recommendations from employed blacks. Elliot (1999) does find that the use of non-white rather than white contacts was linked to lower wages. Falcon (1995) shows that Boston Latinos’ use of personal network lowers their earnings. Green et al. (1999) find that use of informal search methods is disadvantageous for Hispanics - a finding supported by Korenman and Turner (1996) concerning black-white wage differences.

10

Institutional dissimilarity

Most previous research about the importance of job-finding methods and informal hiring practices in job acquisition is about the United States’ labor market, which lacks a national employment agency. Given the largely unregulated character of the U. S. labor market, it seems likely that networks may play a greater role there than in many West- European countries. Some European researchers suggest that regulatory agencies and greater bureaucratization with respect to entry requirements play a greater role in the West European employment system than in the United State (Wegener 1991, De Graaf & Flap 1988, Manwaring 1984; Roper 1988).

In contrast to the United States, Sweden has a nationwide system of public employment agencies, with at least one office in each municipality. The difference in search context between the two countries involves both the extent of the job information available as well as the ease at which this information is obtainable. Employers in Sweden are required, by law to report vacancies to the National Labour-Market Board3. Job information is obtainable from the public employment agencies either through frequently published listings of job openings, continuously updated computer-based vacancy services, or the Internet. Furthermore, practically all unemployed persons visit the local employment agencies. Some of these visits will be related to the claiming of unemployment compensation (as regular visits to the employment agencies are a requirement for obtaining unemployment

11

compensation4), yet, most visits are likely to be associated with some degree of job search. The question is thus if social networks are less important for job finding in Sweden than in the United States? According to those few studies about recruitment and job finding in the Swedish labor market, this does not seem to be the case. Okeke (2001) reports that during 1990-1998 only 20 to 26 percent of “newly employed persons” had found a job through formal job-finding methods. Corresponding figures for “unemployed” individuals who found a job through formal job-finding methods in the same study are between 23 to 40 percent. Ekström’s study (2001) of recruitment behaviour of Swedish employers shows that the informal channels were the main channels for recruiting the workforce (about 65 percent). Korpi (1998) demonstrates that 47 percent of a sample of unemployed Swedes contacted friends, relatives, or acquaintances when seeking work.

12

Data, measures and Empirical set-up

Our analysis is based on data obtained from the Swedish Labour Force Surveys (AKU, Arbetskraftsundersökningar), conducted by Statistics Sweden. The Labor Force Survey (AKU) is based on a sample of individuals, which represent the Swedish adult population (age 16 and above). Questions about job contacts and wages were asked in the January surveys of the AKU 1992-1999, and information in AKU January 1992-1999 consider labour market condition of the pervious years, which is 1991-1998. Pooling January AKU 1992-1999, the data set became large enough to allow studying the role of contacts among immigrants. Systematically excluded from the data are those individuals who worked fewer than 20 hours per week, selfemployed and those who had monthly pre-tax wage less than 5 500 SEK5. The sample is restricted to individuals who under each year; i) were unemployed at the beginning and found a job within the year (3 168 individuals), ii) were “out of labor force” at the beginning and got a job within the year (1 646 individuals), and iii) changed their jobs to new jobs in another workplace (2 073 individuals). In other words, only those respondents who found or changed job in the previous year are included in this analysis (totally 6 887 individuals). Among cases in the whole sample, 153 cases with missing value in some variables are dropped. Deleted cases were a random subsumple of the whole sample. A summary of respondent characteristics is shown in Appendix, tabel A1.

13

Our main independent variable is job-search strategy. Respondents who found/changed jobs during the previous year were asked about the search methods used which had led to their new job. The data, in this way measured retrospectively the single method through which individuals successfully located work. The following search methods were specified: 1) formal methods (governmental employment agencies, private employment agencies, newspaper ads), 2) informal methods (help of friends/relatives or employer made contact with me), 3) direct application to employers or 4) some other method. These methods are represented as a dummy variable, with informal methods as the reference category. Furthermore our data is not sensitive enough to disentangle the degree of informal methods in “direct apply” and “other” methods. Therefore, we will focus on comparison between informal employment procedures and other channels.

In addition to job-search methods, we also control for the following background characteristics: education (measured by years of schooling), potential labour market experience6, gender (dichotomous, with men coded 1), trade union membership (dichotomous, with members coded 1), employed in the public /private sector (dichotomous, with public coded 1) and place of birth (with native-born Swedes as reference group). We also use dummy variables for the years in which the job attainment of concern to us took place (dichotomous, with 1991 coded 1); these

14

dummy variables control the time trend and the impact of the macro-level economic conditions on the job-search process. Income in this analysis is the selfreported pre-tax wages from the respondent’s current job. Wages are measured on a monthly basis and are logged to create a normal distribution.

Unfortunately, our measure only deals with successful job finding techniques used by the respondents. This means that we omit those who looked for jobs but failed to find any. It is plausible to assume that immigrants (particularly those from ONW countries) are overrepresented in this group (see for example Arai & Villhelmson 2001).

We start the next section by giving a brief description of the average differences of various groups in their use of different job-search methods. We then analyze the same problem by estimating a logit equation in order to investigate to what extent each group used informal methods for finding a job. Thereafter, we examine the impact of the macro level economic conditions on the job-search process. Next we employ an OLS Model to analyze what effects informal job-finding methods versus other methods, have on wages, and we estimate wage differences between immigrants and natives, when they used various job-search methods. Finally we repeat the same procedure separately for men and women in our sample.

15

Empirical results

The question that we examine in this section is whether groups with particular characteristics face different opportunities and, consequently, turn out to be more or less prone to use certain job-finding methods. In this context, variables of region of birth, gender, educational background, social group and age are controlled. Table B. 1. in Appendix presents job-finding methods by various backgrounds of respondents. Results indicate that some of these characteristics have an impact on the likelihood of using a certain method. The results also suggest that natives have more success in finding a job through informal methods; 43 percent for natives versus 36 percent for born in NW countries and 34 percent for those born in ONW countries.

A logit equation predicting a dichotomous variable indicating the use of informal methods was also estimated, with results shown in Table 1. There are significant differences between natives and immigrants in using informal job-finding methods. Informal methods tend to be used to a greater extent by natives. Immigrants from NW countries and immigrants from ONW countries found their jobs informally 19 percent and 26 percent less than our reference group, respectively. These results suggest that differences between natives and immigrants in employment rate seem, in part, to be attributable to differences in contacts. The statistically significant

16

coefficients are also for union membership, employment in private sector, genus and years. Women used informal methods for job finding 12 percent less than men. Employed individuals in private sector used informal methods almost 70 percent more than those employed in the public sector7 (about variables for the years see the next section).

Table 1. Determinants of using informal job-finding methods Variables in the Equation

Exp (B)

Constant

.362***

Born in NW

.807*

Born in ONW

.74**

Not union member

1.192***

Employed in Private sector

1.69***

Female

.885**

Year 93

1.31**

Year 94

1.693***

Year 95

1.519***

Year 96

1.581***

Year 97

1.821***

Year 98

1.877***

Year 99

1.693***

Upper secondary school (ref: compulsory school)

.894

University +

.80***

Potential labour market experience

1.0

-2 Log likelihood

8908.083

N

6 734

* denote significance at 10 percent level, ** at 5 percent and *** at one percent.

17

The economic conditions

In this section we examine whether the tightness or looseness of the labor market have an impact on job-finding methods above and beyond the situation of any particular job seeker, or to what extent the effectiveness of various strategies is conditioned by the business cycle. Accordingly, we compare the search strategies, which had led to a job both in times of economic growth and economic recession.

The impact of the macro level economic conditions on the job-search process has been investigated in few studies (Wood 1985, Wial 1991, Licht 1992 and Alon and Stier 1997). The most general argument being that in recessions, informal recruitment method increases.

The time span of this study (1991-1998) facilitates an investigation of the effects of economic condition on the efficiency of job-search methods. The Swedish labor market experienced a dramatic change in the early 1990s. The overall unemployment rate was 3 percent in 1991. By 1993, the unemployment rate had increased to 8.2 percent. Employment ratios fell substantially from 81 percent in 1991 to 70,7 in 1997. Table 2. Unemployment rate and Employment ratio. 1991-1998 Year

1991

1992

1993

1994

1995

1996

1997

1998

Unemployment rate

3.0

5.2

8.2

8.0

7.7

8.1

8.00

6.5

Employment ratio

81.0

77.3

72.6

71.5

72.2

71.6

70.7

71.5

18

The question is if there is a correlation between using of informal job-search methods and the economic conditions? The results in the Table 2. demonstrate that with 1991 as reference year, the odds ratios for using informal methods in all other years are more than one. It means that at time of economic recession (with high unemployment rate and low employment ratio), it is more likely that individuals find a job via social networks, in comparison with other job-finding methods. To be more precise, we measure the correlation between macro level economic conditions and informal job-search method. We obtain the Pearson correlation coefficient for employment ratio for each year and odds ratio for using the informal methods in the same year. The result demonstrates a very strong and statistically significant correlation (- 0.935 and significant at the 0.01 level 2-tailed)8, between economic conditions and utilizing informal job-search methods under period of our investigation9.

Granovetter (1995:159) argues that in such an investigation we must consider the other aspects which may affect using of informal job-search methods under recession, for example whether mass layoffs are geographically concentrated or not. He suggests further that the looser the market, the greater the importance of recruitment strategies by labor buyers relative to behavior of sellers of labor power. Thus in recession it is employers’ recruitment strategies rather than workers’ job search strategies that require study.

19

The impact of informal job-finding method on wages

We specify at the first stage a wage equation including so-called “human capital” variables (education and potential experience), and other variables, such as birthplace, gender, trade union membership, employment in the public/private sector and years in which the job attainment/change took place. Control for “jobsearch methods” are entered in to the equation in model 2.

In table 3 we present regression equations in which we have estimated the effects of the informal job-finding method on (the natural logarithm of) the monthly wages. Employed individuals with ONW immigrant background10 have a monthly wage, which is about 14 percent less than natives when education, potential experience, birthplace, gender, having children, civil status, trade unions membership, employment in the public /private sector are controlled (Model 1).

20

Table 3. The impact of informal job-finding method on (log) monthly wages. OLS Coefficients. (standard errors in parentheses)

Model 1

Immigrant (from ONW countries)

Exp (x) -1

B

Exp (x) -1

B

Exp (x) -1

-.145***

-.135

-.143***

-.133

-.12***

-.11

(.016)

.03*** (.007)

Immigrant ONW x Informal

Model 3

B

(.016)

Informal methods

Model 2

(.020)

.03

.032***

.032

(.007)

-.066**

-.063

(.034) R2adjusted N

.373 .374 .375 6 733 6 733 6 733 *** denote signifaicane at one percent level and ** at five perecent level. All models include controls for: education. potential experience. gender. child. married. trade unions membership. public /private sector employment.

Controls for “informal methods” are summarized in Model 2. Job found through informal methods pay about 3 percent more than other jobs. Controlling for whether the job was obtained through the informal methods does not significantly change the differences in wage between ONW immigrants and natives.

In Model 3, we have added an “immigrants-informal” interaction term to test the proposition that informal job-finding methods have an effect on the wages of immigrants and natives. Inclusion of the interaction term obviously changes the

21

interpretation of the coefficients of the “informal methods” and “immigrant” in the following way: a) The difference in wage between ONW immigrants and natives, who found their jobs through methods other than informal methods, is given by the coefficient of the “immigrant” variable. The coefficient of -0.11 on the “immigrant” variable indicates that the wage of ONW immigrants who found job not through informal methods stood about 11 percent below those of employed natives who found their jobs in a similar way.

b) The difference in wages between natives who found their jobs through personal contacts and natives who did not is given by the coefficient of the “informal methods” dummy variable. Those natives who used informal methods to find a job have a wage which is 3.2 percent higher than those natives who did not. Corresponding (within group) social network effect for immigrants is given by the sum of the coefficients of the “informal methods” variable and “interaction” term. The result indicates that those ONW immigrants who used “informal methods” to find their job have a wage which is about 4 percent lower than those immigrants who used other methods.

c) The difference in wage between natives and ONW immigrants who found their jobs through informal methods is given by the sum of the coefficients of “immigrant” variable and the “interaction” term. The wage of ONW

22

immigrants who found jobs through informal methods stood about 17 percent below those of similar natives who found their jobs through informal methods. In every step (from model 1 to model 3), we have tested whether the addition of a new variable significantly increase R2 above the R2 predicted by a set of variables already in the equation11.

The results indicate that there is a significant difference between the wages of ONW immigrants and natives when we control for “human capital” variables and our other control variables (Model 1). Inclusion of the “informal methods” variable has very little effect on the immigrants versus natives wage differences (Model 1 versus 2). But more importantly, ONW immigrants earned less than natives, because ONW immigrants’ social networks “pay” far less than those of similar natives. There is also evidence of lower payoff to informal methods among ONW immigrants (Model 3).

Finally we estimated coefficients of Model 3 separately for women (Model 3w) and men (Model 3m) in our sample, because some previous studies indicate that women’s networks are less likely to include contacts in possession of valuable job information (see Mardsen and Gorman 2001:476). Table 4 shows the result of these estimations. Model 3w shows only that wage difference between native and immigrant women from ONW countries is not statisticaly sinificant. Using social

23

network increase the wage for nearly 2 percent for all women, because interaction varible is not statistically significant.

But for men, on the other hand we have three statistically significant coefficients. The coefficient of -0.19 on the “immigrant” variable indicates that the wage of ONW immigrant men who found job through job-search methods other than informal methods stood about 19 percent below those of similar employed natives men who found their job in a similar way. Those native men who used their “social networks” have a wage which is 4 percent more than those native men who did not. And finally the wage of ONW immigrant men who found jobs through their social network stood about 26 percent below those of similar native men who found their job through informal methods [(-.19)+ (-.075) = .26.5].

24

Table 4. The impact of informal job-finding method on (log) monthly wages, separately for men and women. OLS Coefficients. (standard errors in parentheses)

Model 3W

B

Exp (x) -1

-.036

-.21***

-.19

(.029)

(.028)

B Immigrant (from ONW countries)

Model 3M

Informal methods

.017* (.010)

Immigrant ONW x Informal

-.03 (.050)

Exp (x) -1

.017

.04*** (.010)

.04

-.078* (.048)

-.75

R2adjusted N

.26 .37 3 372 3 360 *** denote signifaicane at one percent level and * at 10 perecent level. All models include controls for: education. potential experience. gender. child. married. trade unions membership. public /private sector employment.

Our result about the more negative effect of social networks on ONW immigrants’ wage is in accord with Mier and Giloth’s (1985), Falcon’s (1995), Korenman and Turner’s (1996) conclusions that Hispanics in the United States have been hurt by employers’ informal recruitment strategies. Also Grenn et al. (1999) report that Hispanics who rely on informal job-search methods get a poorer pay-off from their job search than those who use formal search methods. In addition MostacciClazavara (1982), found that working-class ethnic ties often led to jobs, but that the lower the average income in one’s ethnic group, the less advantage in income from using one’s same-group contacts, compared to other methods of finding jobs.

25

Summary and Discussion

The results we have reported above are, for the most part, quite consistent with those of prior works on job matching. We find that there is a significant positive correlation between lower level of employment ratio and increasing of recruitment via informal methods, which may be interprets the tighter the labor market, the greater the importance of informal recruitment strategies.

We find further that immigrants are less likely to be able to find their jobs through their social networks, in comparison to natives. If personal contacts may facilitate the transition to employment, less ability of immigrants to use the social networks could explain part of the differences in employment rate between immigrants and natives, in the Swedish labor market. But the differences between natives and immigrants are not limited to differences in employment rates.

The other important aspect is the quality and characteristics of jobs, which are measured here by the wage of employees with the same productivity. Differences in wages between ONW immigrants and natives are not explained simply by a lack of personal contacts to jobs among ONW immigrants, because controlling for jobfinding methods has very little effect on estimated wage difference. In other words the disadvantage of ONW immigrants concerning their wages, are not principally

26

explained by a lack of contacts to jobs. The lower wages of ONW immigrants appear to be explained partly by the fact that jobs found through contacts did not pay as well for ONW immigrants as for natives, that is, informal job search did not lead to better outcomes for ONW immigrants, rather the opposite. Among individuals who found jobs through informal methods, natives had better-paying jobs (20 percent more for all individuals and 28 percent more for men) than ONW immigrants. While the difference between ONW immigrants and natives who found their jobs without using social networks was not so great (12 and 18,5 percent). Hence, ONW immigrants who relied on their social network get a poorer pay-off from their job search than do those who use other search methods, but jobs found using formal methods are better on some dimensions. Finding jobs through social network may be one’s best option, yet the jobs found still be of poor quality by general standards if this is all the group can provide. You cannot get blood from a stone. Our findings suggest although a hypothetical eliminating differences in contacts to jobs may help reduce differences in employment rate between immigrants and natives, but social networks alone would not substantially reduce differences in the wage rates of employed individuals between ONW immigrants and natives.

27

Certain limitations restrict the generality of our findings. Our analysis was done only on individuals who were hired. Only successful cases of job acquisition may not reflect the general picture of the effects of social segregation of ONW immigrants and its consequence on obtaining a job. We could not compare those who found jobs to those who did not, but could only better and worse paid jobs. In addition, the 1990’s were an exceptional period in the recent economic history of Sweden with a dramatically economic downturn. ONW immigrants were hit much harder than natives under this period by higher unemployment rate and lower labour force participation rate (see Integrationsrapport 2003). As our data do not contain the non-working population, we have captured only a selected group of successful ONW immigrants who got a job.

Data in this study, like many other studies about job-search methods and job matching, only record the type of tie used to get a job and do not include any information on individuals’ networks composition. It is not only “networking” per se but also a particular kind of networking that provides individual advantage. Not all personal contacts are alike; some provide much better access to information and influence than others. This demonstrates the need for surveys, which include more comprehensive information on an individual’s network composition and her/his social resources. This would enable researches to yield a clearer account of the process through which network structure affects employment outcomes. It is not adequate to look only at the nature of the tie between job finder and her contact, but that

28

various characteristics of the entire network affect outcome. Researches based on social resources theory (see for example Lin 2001) call attentions to the composition of network and contend that those to whom one can gain access, rather than channels through which access is gained, most directly affect attainment.

29

Notes

1.

Countries defined as NW in this study are: Denmark, Finland, Norway, Island, UK, France, Italy, Germany, Netherlands, Belgium, Luxemburg, "other Western European countries”, Japan, Canada, Australia and the US. The rest of the world is defined as ONW.

2.

Homophily principle states that persons linked by social ties tend to resemble one another in numerous ways. Substantial evidence exists to support this claim for a variety of dimensions such as class, sex and race/ethnicity (Lin 1999).

3.

This does not guarantee that all openings actually are reported to the public employment agencies.

4.

Surveys shows that almost 90 percent of unemployed jobseekers visit the employment agencies (se for example L. Behrenz 1998:27).

5.

See also Le Grand and Szulkin (1999:96) regarding minimum monthly wage.

6.

Bridges and Villemez mention that the use of ”work experience” as a control variable may be problematic if in fact this is in part a proxy for “social capital” presumably including some aspects of social network (1986:579). And Granovetter (1995:149) comments that any cross-sectional analyses may miss the role of personal contacts in building a career. If the benefits of good early jobs found through contacts then translate into later labour market advantage, more variance is in fact attributable to social network than can be captured in cross-section.

7.

This result is in accordance with Mardsen (1996), which reports much lower use of referrals by public sector establishments than by those in private in the U.S.

8.

Pearson correlation coefficient for unemployment rate and odds ratio for using the informal methods in the same year is + 0,892 and significant at the 0.01 level (2-tailed).

9.

In addition the Spearman correlation coefficient for employment ratio for each year and odds ratio for using the informal methods in the same year is -.717

10. (significant at 0.05 level). 11. The difference in wages between immigrants from NW countries and native-born Swedes was samll and not statistically significant. 12. See the Test of Addes Subset of independent varibles in Tabachnick and Fidell (2001:145).

30

References Alon, Sigal, and Haya Stier (1997) "Job Search, Gender, and the Quality of Employment in Israel." Research in Social Stratification and Mobility 15:133-149. Arai, M., and R. Villhelmsson (2001) "Immigrants' and Natives' Unemployment-risk: Productivity Differentials or Discrimination?" Stockholm: FIEF. Behrenz, L. (1998) Essays on the employment service and employers' recruitment behaviour, Lund, Nationalekonomiska institutionen, Lund University. Braddock, Jomills Henry, II, and James M. McPartland (1987) "How Minorities Continue to be Excluded from Equal Employment Opportunities: Research on Labor Market and Institutional Barriers." Journal of Social Issues 43:5-39. Bridges, William P., and Wayne J. Villemez (1986) "Informal Hiring and Income in the Labor Market.", American Sociological Review, 1986, 51, 4, Aug, 574-582. Coverdill, James E (1998) "Personal Contacts and Post-Hire Job Outcomes: Theoretical and Empirical Notes on the Significance of Matching Methods.", Research in Social Stratification and Mobility, 16: 247-269. De Graaf, Nan Dirk, and Hendrik Derk Flap (1988) "'With a Little Help from My Friends': Social Resources as an Explanation of Occupational Status and Income in West Germany, The Netherlands, and the United States.", Social Forces, 67 (2):452-472. Ekström, Erika (2001) Arbetsgivarnas rekryteringsbeteende. Uppsala: Institutet för arbetsmarknadspolitisk utvärdering (IFAU). Falcon, Luis M. (1995) "Social Networks and Employment for Latinos, Blacks, and Whites.", New England Journal of Public Policy, 11(1): 17-28 Fevre, Ralph. 1989. "Informal Practices, Flexible Firms and Private Labour Markets.", Sociology, 23(1): 91-109. Granovetter, Mark. 1995. Getting a job : a study of contacts and careers. Chicago ; London: University of Chicago Press. Granovetter, Mark S. (1974) Getting a job : a study of contacts and careers. Cambridge, Mass. Green, Gary Paul, and Leann M. Tigges (1999) "Racial and ethnic differences in job-search strategies in Atlanta, Boston, and Los Angeles.", Social Science Quarterly, 80 (2): 263-279. Holzer, Harry J. (1987) "Informal Job Search and Black Youth Unemployment.", American Economic Review, 77 (3): 446-453 Holzer, Harry J. (1988) "Search Method Use by Unemployed Youth." Journal of Labor Economics 6 (1):1-20. Integrationsverket (2004) Rapport Integration 2003, Norrköping: Integrationsverket Sweden. Korenman, Sanders, and Susan C. Turner (1996) "Employment contacts and minority-white wage differences.", Industrial Relations, 35 (1): 106-123 Korpi, Tomas (2001) "Good Friends in Bad Times? Social Networks and Job Search among the Unemployed in Sweden.", Acta Sociologica 44 (2): 157-170 leGrand, C. and R. Szulkin (2002). "Permanent Disadvantage or Gradual Integration: Explaining the Immigrant-Native Earnings Gap in Sweden." Labour 16(1): 37-64. le Grand, Carl, and Ryszard Szulkin (1999) Invandrarnas löner i Sverige : betydelsen av vistelsetid, invandrarland och svensk skolgång. Stockholm: Univ. Institutet för social forskning.

31

Loury, Glenn C. (2002) The anatomy of racial inequality. Cambridge, Mass.: Harvard University Press. Manwaring, Tony (1984) "The extended internal labour market." Cambridge Journal of Economics, 8 (2): 161-188. Marsden, Peter V. (1996) "The Staffing Process: Recruitment and Selection Methods." Pp. 133-156 in Organizations In America: Analyzing Their Structures And Human Resource Practices, edited by Arne L. Kalleberg, Knoke, David, Marsden, Peter V.,. CA: Sage. Marsden, Peter V., and Gorman Elizabeth H. (2001) "Social Networks, Job Changes, and Recruitment." Pp. 467-502 in Sourcebook of labor markets : evolving structures and processes, edited by Ivar E. Berg and Arne L. Kalleberg. New York: Kluwer Academic. Marsden, Peter V., and Jeanne S. Hurlbert. (1988) "Social Resources and Mobility Outcomes: A Replication and Extension.", Social Forces 66(4):1038-1059 Marx, Jonathan, and Keven T. Leicht (1992) "Formality of recruitment to 229 jobs: Variations by race, sex and job characteristics." Sociology & Social Research, 76(4):190-197. Mier, Robert, and Robert Giloth (1985) "Hispanic Employment Opportunities: A Case of Internal Labor Markets and Weak-Tied Social Networks.", Social Science Quarterly, 66(2): 296-309 Okeke, Susanna (2001) Arbetsförmedlingens marknadsandelar. Solna: AMS utredningsenhet Arbetsmarknadsstyr. Osterman, Paul (1980) Getting started : the youth labor market. Cambridge, Mass.: Mit P. Petersen, Trond, Ishak Saporta, and Marc-David L. Seidel. 2000. "Offering a Job: Meritocracy and Social Networks.", American Journal of Sociology, 106(3):763-816. Ports, Michelle Harrison (1993) "Trends in Job Search Methods, 1970-92.", Monthly Labor Review, 116 (10): 63-67 Roper, Stephen (1988) "Recruitment Methods and Vacancy Duration." Scottish Journal of Political Economy, 35: 51-64. Tabachnick B, and Fidell Linda S. 2001 Using Multivariate Statistics. Boston, MA: Allyn and Bacon. Wegener, Bernd (1991) "Job Mobility and Social Ties: Social Resources, Prior Job, and Status Attainment.", American Sociological Review, 56(1): 60-71

32

Appendix : Table A 1. Summary of sample characteristics (Percent or Mean)

N Age Years of Education Experience in present job (years) Wage (monthly, SEK) Gender-female (%) Kids < 18 (%) Married(%) Private Sector (%) Union member (%) Employment background were unemployed were “out of labor force” changed to new job

Natives

NW

ONW

6 134

294

306

31.8 11.8 12.8 13 0004

37.2 11.7 18.2 13 609

33.9 12.6 14 12 012

50 37 59.5 68 75

55.8 48.6 64.6 69 76

49.3 52 65.4 58 75

45.1 24 30

53.4 16.7 30

57.2 27.5 15.4

significance test

.00 .00 .00 .00

33

Appendix : Tabel B.1. Job-finding methods by various backgrounds of respondents (means %) Formell Informell Dir. apply Annat Born in Sweden Born in NW Born in ONW

26 32 34

42 37 34

24 23 21

5 6 5

Men Women

25 30

46 39

23 26

6 5

Compulsory school Upper secondary school University +

24 26 35

35 43 39

22 26 19

8 5 8

Unskilled blue-collar Skilled blue-collar White-collar. low level White-collar. middle level White-collar. high level

25 20 34 32 35

44 44 40 39 41

26 29 20 22 18

5 7 6 6 7

Private sector Public sector Age: 16-25 26-32 33-45 46-64

25 33

47 32

23 28

5 7

24 31 30 26

45 38 41 45

27 26 23 18

4 5 5 11

Union member Non union member

28 25

41 46

25 23

6 6

Changed job Out of labour force Unemployed

26 21 32

46 42 40

18 31 25

10 6 3

34