Appendix I: Agency game: How social capital lowers transaction cost.
This simple agency game will show that transaction cost can be decreased by social capital. Note that we defined social capital here as networks, as in the definition of Fukuyama.
Suppose at first that a transaction occurs at the market place, where actor A gives an amount (100) of money to actor B. Actor B has an investment opportunity that will give him a return of 25 percent. He promises actor A to return 110 percent of the money after period t. This will give us the following pay-off matrix.
Player A
Invest Don’t invest Player B Perform (110, 15) (0,0)
Breach (0, 125-OCA) (0,0)
Table 1: payoff matrix of the investment game. OCA denotes opportunity cost of forgone transaction opportunities with player A.
Player B has, once the transaction occurred, an option to breach. As can be seen in the table, he will breach if (125 – opportunity cost > 15). The opportunity cost here consists only of forgone profit from new possible transaction with player A. (Note that here the only relationship between A and B is this investment relationship.) As long as the opportunity cost are not too large, transaction will lead to breach of contract, which will implies that no transaction will take place. This will cost the society 25.
The market place can provide us with an institution (the legal system) that secures this transaction. Of course, the legal system is not free, but paid by taxes. One can image that because of capital tax (tC) the return of the investment decreases. The legal system also requires that the transaction is administrated correctly, which will involve administration cost (a). The legal system however will secure the transaction for player A. The social welfare of the transaction however reduces to 25 – tC – a.
Appendix II: The SPVA Survey
The SPVA survey is a regular intended survey towards immigrants. The intention is to map out the social position and facility usage of immigrants. The survey is only focused on the four largest groups of immigrants: Turks, Moroccans, Surinamese and Antilleans. The survey stated in 1988, was repeated in 1991, 1994, 1998 and had its latest edition in 2002.1 We used the latest edition covering the largest group of immigrants: 2002
The fieldwork is done by Bureau Veldwerk, who sent an interviewer to selected families. This implied that the questions were answered at home. The data were collected by three types of interviews, namely one aimed at the respondent that felt itself the heads of family. This interview contained all questions. The other members of the family (older as twelve) were asked to answer a much shorter list of questions. Finally only one of these other members was asked to answer some additional questions regarding cultural integration. Totally 4199 households were asked to answer these questions, which involved 8321 persons. The questions were asked from February 2002 until April 2003. The response was about 48 percent. More details regarding this research can be found in the codebook of the SPVA survey, available by DANS: (http://www.dans.knaw.nl)
We used the data resulting from the interviews with the head of family, since this list was the only one that included all the variables necessary for our variables.
Appendix III: Composing the bridging and bonding social capital variables
This appendix will show how we composed the bridging and bonding variables. We here used the questions of the SPVA survey that referred to contact with the ethnic minority and native Dutch majority. The following table shows the nine questions from the SPVA survey.
Question from the SPVA Composing the bridging / bonding variable
Do you take care of persons in your land of origin? Yes = bonding, No = bridging
Do you have Dutch friends? Yes = bridging, No = bonding
Do the organizations from which you are a member have mostly Dutch members or mostly Ethnic members?
Mostly Dutch=bridging, Mostly native=bonding, mixed: both
Are there sometimes Dutch friend and/or neighbours visiting?
Often = bridging, Never = bonding
Do you spent time with native Dutch during spare time Often = bridging, Never = bonding Do you have more contacts with ethnic group or with the
native Dutch
More contact with Dutch=bridging, Mostly ethnic = bonding
Was your partner born in the Netherlands, or in your land of origin?
Netherlands = bridging, Foreign = Bonding
Would you mind when your child had Dutch friends? Not at all or totally not at all = bridging, Yes or absolutely yes = bonding
Would you mind when your child had a Dutch partner? Not at all or totally not at all = bridging, Yes or absolutely yes = bonding
Appendix IV: Measurement of the size of the network
The survey had some questions referring towards the people surrounding the respondents. These were:
1. Do you have a partner?
2. Are you living together with others? (Partner, friends, children?) 3. Do you visit religious meetings at least once a month?
4. Are you a member of one or multiple voluntarily organizations?
We combined the answers of these questions into one network size variable. We calculated these by increasing the value of the network size variable with one for each positive answer towards the first three questions. Furthermore the number of voluntarily organizations was added to this network size variable. So suppose that a respondent has a partner and lives with his family, visit a church regularly and is member of both a soccer club and a political organization, the value of the network size variable becomes 5.
Appendix V: Cultural Integration
The survey asked many different questions regarding cultural integration. We used only included the opinions of the respondents regarding some standardized propositions since these clearly referred to opinions instead of the real situation of the immigrant. Furthermore since the answers had to be given on a scale with the answer possibilities of I agree totally, I agree, I’m do not agree nor disagree, I don’t agree, I totally do not agree, we could easily use these answers into factor analysis.
The following questions are used:
Question from the SPVA Kind of Cultural
Integration involved It’s better that at the end, the man decides regarding large purchases. Emancipation It’s the best that the woman takes responsibilities for the house holding. Emancipation Responsibilities regarding money are in best hand with the man. Emancipation
Education is for boys more important as for girls. Emancipation
A woman can be happy even without children. Emancipation
Earning an own income is for boys more important as it is for girls. Emancipation When getting children, a woman has to quite working definitely. Emancipation
Man and woman are allowed to life together unmarried. Liberal Values
Families with young children are not allowed to go on a divorce.
Married and two children is still the best form of family. Liberal Values It’s the best that children are living at home until they are getting married. Liberal Values For a seventeen years old daughter it is allowed to live on her own. Liberal Values For a seventeen years old son it is allowed to live on his own. Liberal Values
A child should decide it selves who to marry. Liberal Values
Question from the SPVA Kind of Cultural Integration involved Someone who hasn’t long to life anymore and lot’s of pain should be allowed to decide it
selves regarding ending its life.
Liberal Values
It is wrong when your daughter wants to marry someone with another religion*. Liberal Values It is wrong when your son wants to marry someone with another religion*. Liberal Values It is wrong that in the Netherlands the influence of religion is decreasing largely.* Liberal Values The school of children should be fitting into our religion.* Liberal Values Aged parents should be living together with their children. Influence of family
Children should take care of their sick parents. Influence of family
You should always respect your parents, even when they do not earn this because of their behaviour.
Influence of family
Relations within the family are very close. Influence of family
When you’re having problems, family should help. Influence of family
It’s normal that parents give money to their grown-up children to support large expenses.
Influence of family
Appendix VI: Characteristics of the respondents by ethnic group
Some characteristics of the respondents by ethnic group are showed in the following tables. It follows that the differences between the diverse ethnic groups are quite large.
Highest completed education of the ethnic groups Total
Max. primarily1 Lower secondarily education2 Higher secondarily education / lower professional education3 Higher professional education4
Ethnic Group Turks 49,4% 22,8% 20,8% 7,1% 100,0%
Moroccans 55,9% 13,8% 21,7% 8,6% 100,0%
Surinamese 24,5% 29,4% 31,3% 14,8% 100,0%
Antilleans 19,2% 29,1% 31,3% 20,4% 100,0%
Total 38,4% 23,6% 25,9% 12,2% 100,0%
Table 1: Attained education by subgroup 1: In Dutch terms: Basisonderwijs
2: In Dutch terms: VBO / MAVO / VMBO 3: In Dutch terms: HAVO / VWO / MBO 4: In Dutch terms: HBO of university
Proficiency of Dutch of the ethnic groups
Total Percentage with lowest score Percentage with highest score
Ethnic Group Turks 16,8% 24,5% 100,0%
Moroccans 7,4% 42,8% 100,0%
Surinamese ,3% 83,9% 100,0%
Antilleans ,6% 68,8% 100,0%
Total 6,8% 53,8% 100,0%
Table 2: Proficiency of Dutch of the four ethnic groups.
Percentage self employed
Reality Our data
Ethnic Group Turks 6% 10,8%
Moroccans 3% 7,0%
Surinamese 4% 5,4%
Antilleans 3% 5,1%
Total 6,4%
Figure 1: Details regarding deviation of income of Turks.
Figure 2: Details regarding deviation of income of Moroccans.
Figure 3: Details regarding deviation of income of Surinamese.
Appendix VII: Characteristics of the respondents by level of education and job position
Highest received
certificate Mean N Std. Deviation Elementary 2,5678 2168 1,06948
Lower 3,5437 470 ,46635
Medium 3,6321 573 ,36612
Scientific 3,6802 359 ,31985
Total 2,9791 3570 1,00865
Table 1: Average level of proficiency for the levels of attained education.
Level of position Mean N Std. Deviation
Elementary 2,7437 349 1,01857 Lower 3,1553 756 ,86943 Medium 3,4380 699 ,62131 Higher 3,5857 305 ,50743 Scientific. 3,6767 94 ,32100 Total 3,2617 2205 ,81458
Appendix VIII: Reasons to change of job Reason to change job % of total respondents Percentage of cases Average income Average bridging Average bonding Average size of the network The job did not
please me
4.6 15.8 1408.16 4.17 2.33 2.17
I did not like the colleagues
0.9 3.3 1473.25 4.50 2.38 1.95
Opportunity to make a promotion
6.1 21.0 1669.71 4.72 1.83 2.23
The job was to far away
1.1 3.8 1320.72 3.85 2.26 2.17
I got too les paid 3.3 11.6 1496.83 4.14 2.50 2.17
Because of movement
1.0 3.4 1318.05 4.52 2.02 2.01
The job ended 4.2 14.4 1442.67 4.27 2.34 2.02
Other 7.7 26.7 1458.53 4.57 2.10 2.22
Total 28.8 100.0 1420,79 4.17 2.34 2.16
Appendix IX: Some characteristics of the diverse religions Race Highest achieved certificate foreign Highest achieved certificate NL Current level of job position Average Income Average Prof-iciency Time spent in the NL Bridging Bonding Size Network Man Mean 1,26 1,74 2,52 1537,67 2,9311 ,4918 3,7673 2,8200 2,2597 N 2456 2288 1564 1111 2359 2162 2456 2456 2456 Std. Deviation 1,169 1,059 1,052 810,063 1,00725 ,21823 2,03657 2,12676 1,08729 Woman Mean 1,13 1,76 2,63 1176,91 3,1086 ,4743 4,0923 2,2328 1,7822 N 1560 1435 725 512 1491 1311 1560 1560 1560 Std. Deviation 1,101 1,040 1,027 489,480 ,97908 ,23644 1,82169 1,92471 1,03667 Total Mean 1,21 1,75 2,56 1423,84 2,9998 ,4852 3,8936 2,5919 2,0741 N 4016 3723 2290 1623 3850 3473 4016 4016 4016 Std. Deviation 1,145 1,051 1,045 743,400 1,00006 ,22541 1,96206 2,07026 1,09286
Table 1: Some characteristics by race of the respondents from the SPVA survey
Given are the average values of achieved education, income, job position, proficiency, time spent in the Netherlands, bridging and bonding social capital and size of the Network for male and female. The given differences of average education in the Netherlands are not significant, the given differences of time spent in the Netherlands and current job position are significant at 5 percent level and the other given different average values are significant at 1 percent level.
Religion Mean N Std. Deviation
None 1418,70 447 941,422 Hinduism 1238,43 214 712,640 Islam 1157,38 1370 598,912 Christian 1198,65 769 632,318 Others 1168,35 59 548,195 Total 1215,62 2859 685,374
Religion Average proficiency Highest received certificate in the Netherlands Highest received certificate abroad No religion Mean 3,5052 2,40 1,12 N 503 467 519 Std. Deviation ,59550 1,197 1,212 Hinduism Mean 3,4313 1,83 1,19 N 239 237 251 Std. Deviation ,66845 ,992 ,993 Islam Mean 2,5907 1,47 1,05 N 1712 1681 1797 Std. Deviation 1,08581 ,867 1,127 Christian Mean 3,5168 2,02 1,57 N 880 827 908 Std. Deviation ,52452 1,137 1,106 Other Mean 3,3461 1,77 1,63 N 61 55 64 Std. Deviation ,74519 1,031 1,163 Total Mean 3,0389 1,77 1,21 N 3394 3267 3538 Std. Deviation ,98213 1,061 1,148
Table 3: Average level of education and proficiency for the different religions
Ethnic group
Total Turks Moroccans Surinamers Antilleans
Religion No religion Count 57 23 236 204 520
Appendix X: Correlation of the included variables
The following table shows the correlation between the different, in the regression included, variables. We include here the complete population of the SPVA survey.
INCOME MAXEDUF MAXEDUNL NETSIZE BRIDGING BONDING PROF RACE TIMENL LIBER HEALTH FAMCLOSE
