an .4-«1 ,‘ A Joann :- ‘- ".79 , 313‘;- cw. .. , 33')" ‘53: ‘5'“ " _ 3M." - 59211:; 35%?132 “‘1‘ a. {$125- ‘E'vv. {1:71 ‘2‘."3‘n 3 «a. , , x? , am: an; “ um , \ 'IO ... . .u ,0» ”>0 .0" an L's a. .. ”w 2 ' ~ . .e. 1-“ I» a ‘ ultlr‘.~‘-"‘ v‘ - ‘ x, 4.. .5 tht r I, r? 5' .i‘i I & 'l' 334%? ‘i‘ .133 j. ,1: 3 s I. ‘1‘“ .5 . . ’1’, 100? This is to certify that the dissertation entitled MOVING TO A WELFARE STATE: A COMPARISON OF ECONOMIC MOBILITY OF SURINAMESE IN AMSTERDAM AND PUERTO RICANS IN NEW YORK CITY presented by KARIJN G. NIJHOFF has been accepted towards fulfillment of the requirements for the PhD degree in sociology gwfl, (0.4L Major Professors Signature December 15, 2006 Date MSU is an Affirmative Action/Equal Opportunity Institution LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATEDUE DAIEDUE DATEDUE 6/07 p:/ClRC/DateDue.indd-p.1 5'1th MOVING TO A WELFARE STATE: A COMPARISON OF ECONOMIC MOBILITY OF SURINAMESE IN AMSTERDAM AND PUERTO RICANS IN NEW YORK CITY By Karijn G. Nijhoff A DISSERTATION Submitted to Michigan State University In partial fulfillment of the requirements For the degree of PHD Sociology 2006 SWIMMI- imigm 1 en tie labor marl trips in M fit}. the [hit Bilih I Situation CUP.“ Bel-'1 comps . came from dj colonial linI Ml Similar man: IaCiUtir memPIOE‘mc ABSTRACT MOVING TO A WELFARE STATE: A COMPARISON OF ECONOMIC MOBILITY OF SURINAMESE IN AMSTERDAM AND PUERTO RICANS IN NEW YORK CITY By Karijn G. Nijhoff This dissertation compares the impact Of the different contexts Of reception on immigrant economic mobility. Three contexts of reception are defined: the ethnic group, the labor market, and the welfare state context of reception. The analysis focuses on two groups in two different labor markets of two welfare states: Puerto Ricans in New York City, the United States and Surinamese in Amsterdam, the Netherlands. Both groups were citizens of the country they arrived in. This makes their situation comparable, as it excludes them from restrictions imposed by immigration laws. Both groups are also non-white, have a large percentage of female headed households, came from diverse origins in their countries of origin, and came from countries with a colonial link to the countries of destination. The labor markets they entered also share many similarities, including the impact of economic restructuring and the decline of manufacturing occupations. There are differences in the labor markets, as in unemployment levels in the cities, on job growth, and on percentages of immigrants in each city. However, the overall picture indicates relatively similarity. The main difference exists in the welfare state: the United States is defined as a liberal welfare state, with low benefits and provisions, the Netherlands is a social-democratic welfare state, and benefits and provisions are generous. 1 am}; at“; context 0 at: Impact on c Ii.” 'm'n. 9; VuknhllliltIOI utterances Li. titled to the 0 an 'ethmc Ct‘ ‘fin-‘l. V. Ll'igsisblonf acrtss COUH' I analyze these aspects in detail, to provide a model for the possible impacts of each context of reception. The main results reflect that the welfare state most likely has an impact on education levels of the groups, while labor market characteristics and discrimination explain most of the lack of occupational mobility of the groups. The statistical analysis of the dissertation encountered problems: the sample sizes were differences across groups, which distorts significance levels. The second problem is linked to the measurement model and the usage of factor analysis. The measurement of an ‘ethnic context of reception’ is problematic and it might be advisable to use separate variables instead of a combined factor to include group characteristics in a model. The conclusions elaborate on alternative approaches to the analysis of occupational mobility across countries. iv Voor pap Acknowlet people. on U and in} com Alan Rad} tl committee :l latte: thci tould- not In }0LL for all It ."fl 1 ~ Acknowledgements This dissertation has been completed through the help of an enormous amount of people, on two sides of the ocean. First of all, I want to thank my advisor Steve Gold and my committee members, Leslie Moch, Alesia Montgomery, Alex Rodrigues, and Alan Rudy (in alphabetical order!). I have been severely lucky to be able to have a committee that included people from a variety of backgrounds and specializations. Without their support and patience, the process towards the end of the dissertation would not have passed as pleasantly. Tammy Spangler needs a very special thank you, for all her help, assistance, and for her friendship. John Schweitzer and Stan Kaplowitz were very kind and helpful when the statistical part of the dissertation needed refining. In the Netherlands, a multitude of people have assisted me in my search for information. I am very gratefiil for all the help, and I am still amazed by the ease of which it was to contact people and meet with them. My special thanks go to Jan Rath, Leo Lucassen, Mies van Niekerk, and Jaco Dagevos. I also had important email- sessions with Ellie Vasta. This list is by no means exhaustive, I have had so many positive interactions with so many people over the years. Thank you all. My colleges and friends at MSU should know they contributed without even being mentioned: the support system among graduate students was of a value that cannot be expressed. And then all the others, families and friends - my global network of inspiration and suggestion! Each and every one of them deserves a separate mention. And of course, I have to thank Brian for his own way of supporting me. 1. lNTRt Contexts ol Ibeoretica Research ( Definitions Structure ( 2. SURll ”The“ Suri Earl} er: lncreasa image lite role lfldi'l’endet Illegal ll! '— lit-gram The POSIIIII labor M, POWQ Edi-Kalli) HllilSm: trEilittl} In some P051. 1 PUER A backgrm . ”Icy-rang] I 1J‘éhlallot Pu . "to RIC; Ho II- “Eli‘s ‘t ePI’SIIIU 981110“ ( ni'l'l‘trtl 'r‘ll)‘ Table of Contents 1. INTRODUCTION Contexts of Reception Theoretical Perspectives Research Questions Definitions and Limitations Structure of the Dissertation 2. SURINAMESE IN THE NETHERLANDS “The” Surinamese Early emigration Increased emigration Images of the new country The role of the state Independence Illegality Migration to Amsterdam The position of the Surinamese in 1982 Labor Market Position Poverty Education levels Housing Family composition Some Post-1983 Remarks 3. PUERTO RICANS ON THE MAINLAND A background to migration Migration to the mainland Migration to New York City Puerto Ricans in New York Housing The Position of Puerto Ricans in New York in the 19803 Position on the labor market Unemployment and labor force participation vi 10 13 15 15 16 18 23 23 24 27 28 29 3O 34 35 37 40 42 48 51 54 56 58 60 60 61 Educator. lncome £1th :01 Changes in I. MlCSRiE lnternation: lit—gratis: The Sin Comm; litigation Minot: immigran llobihtx 2 Economic n Detintttor Internation. Compam Factors (locum (OHCIUSlons 5- DISCR Race and Race and aim OI Elfin" ‘ b :d‘]‘ 6- WELF, The [menu “6 . 's'lns e T} on] [M'gratim F'fare S' ilfare 8 “me Can Education Income Family composition Changes in Position since 1980 4. MIGRATION AND MOBILITY International Migration Migration Theory The Start of Migration Continuing Migration Migration and Natives’ Opportunities Migration and Wage-Parity Immigrant Mobility Theories Mobility and Labor Market Context Economic mobility Definitions International Economic Mobility Comparative Differences: Mobility Regimes Factors on Mobility Occupational Scales Conclusions 5. DISCRIMINATION AND SEGREGATION Race and Ethnicity in the US Race and Ethnicity in the Netherlands Racism or Ethnicism? Discrimination compared Segregation 6. WELFARE STATES The Emergence of Welfare States Welfare State Regimes Origins The Typology Immigration and the Welfare State Welfare State and Mobility Welfare States and Urban Areas Some Cautions vii 62 63 65 66 72 72 73 73 77 80 81 85 88 92 93 95 96 98 1 00 104 106 107 110 113 115 118 125 125 127 127 129 130 132 134 135 l’nited Stat D1t‘fctsnc Dariarcnc Empia . 717””: Shaun 5 Conclusion 7. GROU Position in ‘ The Cities l Migrants ECOTIUIT; l: Ill urr Labor .' GOI SHIT, Roofing Compara‘. United States and the Netherlands Compared Differences in Typology Differences in Ideology Emphasis on Work Regulation of Work Conclusions 7. GROUP AND CONTEXT COMPARISONS Position in the 1980s The Cities Compared Migrants in the Cities Economies of the Cities U-Tum in wages Labor Market Participation and Unemployment Government Measures Housing Comparative Conclusions of the Cities The Groups Compared Migration Type Diversity of the Groups Education levels Self-Employment Female Headed Households Time of Migration Out-migration Language Comparative Conclusions of the Groups Welfare States 8. METHODS AND DATA Datasets The SPVA in the Netherlands Data Specification The US. Census - IPUMS Variable Selection — Ethnic Group Context of Reception Additional Measures Economic Mobility Methodology Comparing Time Periods viii 138 138 139 140 141 142 145 145 147 147 149 152 153 154 154 156 156 157 158 160 162 163 165 165 166 167 168 171 171 172 172 173 174 175 175 178 178 Computing The llodel 9. THEAh Descriptiie S length of 5 Education I Female He. Conclusion Step 1: Chan Education 1 Cnetnplotr Occupattor Frequent Significant Changes ll‘. Changes in Step 2: The 1 Parameter 1 Cami Ana Alternative .5 m.CONCL ConClusions l llelfate 31;, “Eamon a 'dUCalionf MethodOlogi‘ ,ml-‘lt? Siz mills of [h .thsu Elllnlc GIOI Cunation DISCUSSIOH AppENDlX 1 Comparing Contexts The Model 9. THE ANALYSIS AND RESULTS Descriptive Statistics on the Group Characteristics Length of Stay Education Levels Female Headed Households Conclusions on the Numbers Step 1: Changes over Time Education Levels Unemployment rates Occupational Positions Frequencies in Occupational Categories Significance of the Changes Changes in Education Levels and Occupational Categories Changes in Educational Levels and Employment Step 2: The Immrtance of Context Parameter Estimation Factor Analysis Alternative Approach? 10. CONCLUSIONS AND DISCUSSION Conclusions on the Results Welfare states and Economic Mobility Migration and Mobility Education, Unemployment, and Occupation Methodological Issues Sample Size Limits of the Measures Welfare State and Labor Market Contexts of Reception Ethnic Group Context of Reception Occupational Mobility or Position Discussion APPENDIX 1: SURINAMESE EDUCATION SYSTEM APPENDIX 2: DATA COMPUTATIONS ix 181 183 186 186 187 190 194 195 196 197 199 201 202 205 207 209 211 212 213 215 216 216 225 228 230 231 231 232 233 233 235 236 240 241 Amsterdam loh trout} Ethnic Grou Age of pit; ligation Education Female he. The ll elfare The Dependt A categnrt Summar} APPENDlX Step One: A Results of Educattt Employ OCCUPaI‘ttti RRUIIS from Education EmPIOIt REFERENC Amsterdam Job grth in Amsterdam Ethnic Group Context of Reception Recodes Age of population Migration year (StayCensus) Education (MaxEduc) Female headed households (FemHead) The Welfare State: Public Health Expenditures The Dependent Variable: Occupation Categories A category ‘unemployment’ Summary APPENDIX 3: STEPS IN THE ANALYSIS Step One: ANOVA Results of the one way-ANOVA Education Employment Occupation Results from the two-way ANOVA Education Employment Step Two: Context REFERENCES 241 241 241 241 242 243 244 245 245 247 247 249 249 249 249 250 250 251 251 252 253 257 lthleI-B E lahleI—l L lahleI-S F percent Table?!) lr TahleI-l P TableS-I P Tahle3-3 L IIbEeS-I, L‘ percent .anteS-S [g agesZS TabieS-b ft Table3-7 P Tablet-l C Tablet-1 C Tablet] 1, able 7.: J, Table 7-3 L lacleLl E iatlelj p Tahle7.6 I Table 7.7 p Tmal‘l‘cnd 5°?" 3-1, I. ‘35s 9-1.5 Table 93.1) Table 9_3 9 lahle94' T :3le 95. (i table 96 C Table 9_7 T latte -. Table II: 1 Tab e r: : T , s ”we A2 «I . List of Tables Table 2-1: Surinamese migration to the Netherlands 1964-1972 ................................ 21 Table 2-2: Net migration rates and the total number of Surinamese in the Netherlands 1973-2003. ........................................................................................................... 27 Table 2-3: Ethnic background of Surinamese in the Netherlands, 1982 ..................... 30 Table 2-4: Labor market participation working population, in percentages. .............. 31 Table 2-5: Educational level at point of migration, per migration period (in percentages), migrants 13 years of age and older ................................................ 36 Table 2-6: Index of dissimilarity for Surinamese in Amsterdam 1973-1981 .............. 40 Table 3-1: Puerto Rican migration to the mainland, 1900-1970 ................................. 54 Table 3-2: Puerto Ricans in New York City, 1920-1970 ............................................ 56 Table 3-3: Labor market participation (in percentages) .............................................. 61 Table 3-4: Unemployment rates for New York City and for Puerto Ricans, 1980, in percentages ........................................................................................................... 61 Table 3-5: Educational attainment in New York City, 1980 (percentage of persons ages 25 years and older) ....................................................................................... 62 Table 3-6: Mean household incomes in 1980 .............................................................. 64 Table 3-7: Puerto Ricans in New York, 1980-2000 .................................................... 67 Table 4-]: Occupational categorizations ................................................................... 102 Table 6-1: Gini-coefficients, mid-19805 to years around 2000 ................................. 139 Table 7-1: Immigrants in the Cities ........................................................................... 149 Table 7-2: Job growth percentages ............................................................................ 151 Table 7-3: Unemployment levels and labor market participation ............................. 154 Table 7-4: Education categories ................................................................................ 161 Table 7-5: Public social expenditure as a percentage of GDP ................................... 169 Table 7-6: Taxes on the average production worker as a percentage of labor cost... 169 Table 7-7: Percentage public expenditures as a percentage of total expenditures (see appendix 2 for calculations), on health care expenditures. ................................ 170 Table 8-1: Occupational categorizations ................................................................... 177 Table 9-1: Educational levels in the early 19905 and in 2000 (2002), in percentages ........................................................................................................................... 194 Table 9-2: Percentages F emale-Headed Households ................................................ 195 Table 9-3: 95% Confidence intervals Surinamese and Puerto Ricans, over time. 197 Table 9-4: Test of differences in the changes of education levels ............................. 199 Table 9-5: General unemployment rates in each city (from: Table 7-3, chapter 7) .. 199 Table 9-6: Occupational Levels ................................................................................. 201 Table 9-7: Test of difference in changes of occupational categories ........................ 206 Table A2 1: Employment growth Amsterdam 1987-1988 ........................................ 241 Table A2 2: Employment growth Amsterdam 2001-2002 ........................................ 241 Table A2 3: Year of immigration: ............................................................................. 242 Table A2 4: Years in the United States, intervalled (1990) ....................................... 242 Table A2 5: Categories of length of stay in the 1991 SPVA ..................................... 243 xi Table A2 6: Categories of length of stay in the 2002 SPVA ..................................... 243 Table A2 7: Education levels Census: ....................................................................... 244 Table A2 8: Education levels SPVA 1991: from Maxdipl (highest diploma obtained) ........................................................................................................................... 244 Table A2 9: Education levels SPVA 2002: from Maxdipl ........................................ 244 Table A2 10: Total and public expenditure of health (US dollars calculated using PPPs) .................................................................................................................. 245 Table A2 11: Percentage public health expenditures as a percentage of total health expenditures ....................................................................................................... 245 Table A2 12: Occupational Categories: ..................................................................... 246 Table A2 13: Measures used in the dissertation: ....................................................... 248 Table A3 1: One-way ANOVA: changes in educational levels for Surinamese in Amsterdam 1991-2002: ..................................................................................... 249 Table A3 2: One-way ANOVA: changes in educational levels Puerto Ricans 1990- 2000 ................................................................................................................... 250 Table A3 3: Confidence Interval for ‘unemployment’, both groups over time. ........ 250 Table A3 4: Distribution of Surinamese and Puerto Ricans per occupational category, over time. ........................................................................................................... 251 Table A3 5: Tests of Between-Subj ects Effects: ....................................................... 252 Table A3 6: Total Variance Explained Ethnic Group Factor .................................... 254 Table A3 7: Total Variance Explained Labor Market ............................................... 254 Table A3 8: Total Variance Explained Welfare State ............................................... 254 Table A3 9: Factor loadings for Ethnic Group Context of Reception: ...................... 255 Table A3 10: Total Variance Explained for the factor Ethnic group context of reception, using ULS ......................................................................................... 25 5 Table A3 11: Loadings using ULS ............................................................................. 255 Table A3 12: Model Summary: estimate of R Square ............................................... 255 Table A3 13: Coefficients(a) of the regression .......................................................... 256 xii List of Figures Figure 1-1 Contexts of Reception .................................................................................. 6 Figure 8-1: The Complete Model .............................................................................. 185 Figure 9-1: Length of stay, Surinamese in Amsterdam, 1991 and 2002 ................... 187 Figure 9-2: Years in the United States in 1990 and 2000 .......................................... 189 Figure 9-3: Education levels in 1991 and 2002: ........................................................ 19] Figure 9-4: Education levels of Puerto Ricans in New York City ............................ 193 Figure 9-5: Changes in mean education per group .................................................... 198 Figure 9-6: Changes in unemployment levels ........................................................... 201 Figure 9-7: Changes in occupational position Puerto Ricans, 1990 to 2000: ............ 202 Figure 9-8: Occupational changes Surinamese, 1991 to 2002 .................................. 204 Figure 9-9: Changes in mean occupational level, per group ..................................... 206 Figure 9-10: ‘Average’ occupation level per education level over time, Surinamese in Amsterdam ......................................................................................................... 208 Figure 9-11: ‘Average’ occupation level per education level over time, Puerto Ricans in New York City ............................................................................................... 209 Figure 9-12: Unemployment levels per educational category, Surinamese .............. 210 Figure 9-13: Unemployment levels per educational group, Puerto Ricans ............... 211 Figure 10-1: Group, education, and changes in unemployment percentages ............ 218 Figure 10-2: Group, educational level, and change over time in occupational levels ........................................................................................................................... 222 Figure A3 1: Educational levels per group, over time ............................................... 252 Figure A3 2: Graph of the ‘mean level’ of unemployment per group per year ......... 253 xiii 1, Intr AIter th ‘tlhile glob; neu forms . independen end coioniz Email} co ’Jte global c colonizatro and faster I no o‘elfar $0cral righ; With ti fairly fecer plim0rd;&; IIIOSC Who Slaflfid 38 a Saying but 5513538th Shit; Hallo non‘i‘ntttl 1990. 5:. L 1 . Introduction After the Second World War, a ‘new’ period in global history seemed to develop. While globalization started far earlier than the 20‘h century, the ‘new’ period brought new forms of globalization and interconnectedness. First, decolonization and independence of numerous ‘nation-states’ redefined relationships between colonizers and colonized, between ‘periphery, semi-periphery, and core’. Second, and at least partially consequential, migration flows changed in origins and in destinations. Third, the global economy spread its ties even further than in the times of exploration and colonization. Global inequalities became (even more) visible through the increased and faster communication-systems and news liaisons. Finally, nation-states evolved into welfare states, where citizens did not just have civil or political rights, but where social rights emerged. With these changes, different debates emerged. First, the nation-state, though fairly recent in its existence, provided a new ideology in which belonging -in a primordial sense- was defined in narrow terms. With increased migration, many of those who ‘belonged’ felt that others should be excluded. The nation-state ideology started as an inclusionary movement where not just the rich and the noble had a saying, but where the people became central in the legitimation of government. When established, and when expanded to provide for its citizens through a form of welfare state, nation-state ideology became exclusionary. Immigrants where (often) seen as non-entitled to the rights of that nation-state (Anderson 1991; Marshall 1965; Smith 1991). The we Cold it or one: of u: made root: western co the fear of Inc-ch U137: socialist marital. Tl Sizes do: Sit-Elan lQ‘ lt‘klil} immigrazt 1mm: gran $036 “he! States. ltn ch 5‘ USSlQr marks? ' L“ IS The welfare state developed in times of economic growth in the context of the Cold War. These two historical aspects contributed to the foundation for different types of welfare states, based on different ideological assumptions. Economic grth made room for the possibility of providing for an array of benefits in different western countries. The Cold War had more complex consequences. On the one hand, the fear of a ‘revolution’ created state interference in different domains, to avoid too much unrest among the less-well off. On the other hand, the label ‘communist’ or ‘socialist’ had a bad name in some western welfare state and enhanced the view that people needed to provide for themselves, through active participation in the labor market. Thus, while ideological backgrounds run deeper, the development of welfare states diverged as much as similarities existed (Braverman 1974; Polanyi 1944; Swaan 1996). Today, the benefits and provisions of the welfare state are seen as ‘threatened’ by immigration and globalization (combined with a downturn in economic growth). Immigrants are regarded as ‘abusive to the system’ and welfare is seen as a ‘trap’ for those who are entitled to the benefits. While this ideology is stronger in the United States, linked to the stronger ideology of self-sufficiency in this country, the discussion is not unknown in the Netherlands. Capitalism has ‘won’ and the free- market is now without a competitive economic system. Another discussion has a more positive view on the possibilities that the welfare state offers, and sees a strong or well-developed welfare state as a vehicle for economic mobility. With globalization and immigration, there are voices that claim that welfare states cannot afford to be expansive. On the other hand, it could be argued that the long-term consequences of a on. Boo"LS 1““ lhc main rest economic posrttor statistical compo?! Net erlands and l: on Puerto Rican n integrated Puhl 1c the from lndll tdu. Census, lPL'hlS u stops are memo that the informant needs of the res-ea United States. the 5W) Mtuie 1’: and lIttltZation (ii for largest mint): Compli‘hensite, it 3Spots, and inter 1” Order to let soup are Selectcc 19933an T consequences of an elaborate welfare state pay itself back (Amersfoort and Penninx 1998', Borjas 1999a; Borjas 1999b; Hiebert 2002). The main research question is “How do differences in welfare states affect economic positions for migrant minority groups?” I explore this question through a statistical comparison of economic positions of Surinamese in Amsterdam, the Netherlands and Puerto Ricans in New York City over a ten-year time period. Data on Puerto Rican migration are obtained through the University of Minnesota's Integrated Public Use Microdata Series (IPUMS). IPUMS provides data for public use from individual-level samples of the United States population drawn from the Census. IPUMS is preferable to other public data of the census since national origin groups are presented separately. Another motivation for using data from IPUMS is that the information is easy to access and tabulations can be created for the specific needs of the researcher. Although the Netherlands does not have a Census as the United States, there is extensive data available on the Surinamese population. The survey Sociale Positie en Voorzieningengebruik Allochtonen (SPVA — Social Position and Utilization of Provisions by Allochtones) is a longitudinal survey held among the four largest minority groups in the Netherlands (Dagevos 1998). It is a comprehensive, in-depth survey on education, labor market position, social-cultural aspects, and interactions with autochthones. In order to keep time elements as comparable as possible, the time-points for each group are selected as close as possible: the SPVA of 1991 and 2002, the Census of 1990 and 2000. This is a fairly short time span to examine long-term prospects, lim‘lfisl b} dd: $2185 m NO I The 51.11151 pJSTIEOIlS 0in time are coml anahsts esplt hhiii’l this trottdes a m: participation : minonites ant improiement the changes it dlSCUmlnalittl research Rec: Netherlands, 1 Acceptance a: Contexts Tilt? Unite limited by data limitations. The results give an indication of the influence of welfare states in two contemporary economies. The statistical analysis is done in two parts: first, the changes in economic positions of Puerto Ricans in New York City and Surinamese in Amsterdam over time are compared through analysis of variance (ANOVA). The second part of the analysis explores combinations of factors and their impacts. With this dissertation, I hope to show that a more developed welfare state not only provides a minimum standard of living, but also can affect long-term economic participation for minority groups. The welfare state can create programs that aid minorities and immigrants in the improvement of their socio-economic position. The improvement of occupational positions for the two groups is used as an indicator for the changes in socio-economic well-being in general. Questions of identity, discrimination, or integration outside the economic are beyond the scope of my research. Recent events in France, and recent discussions on integration in the Netherlands, show that economic security is not a guarantee for an optimal situation. Acceptance and participation go beyond an economic minimum standard of living. Contexts of Reception The United States and the Netherlands are the focus in this dissertation. The welfare state is part of what can be labeled the ‘govemment context of reception’: it is one of the aspects that governments use to control immigration within their borders. While I analyze the impact of the welfare state on the economic mobility of immigrants, other factors that are linked to economic mobility cannot be excluded. As a framework I use the contexts of reception as defined by Portes and Rumbaut (1996). Fortes 33 gmgmmeri‘. In this 3 can be 3D 1“ m0 gTOUPS the full ran, tn Benefits 2 The rec: aiotd too In Netherland> necessan tc similarities important!) cities hate I manufacrun tout lert pr, pan 0i [he c Portes and Rumbaut (1996) hypothesize three major contexts of reception: the government, the society, and the ethnic community. In this dissertation I focus on the government context of reception. Welfare states can be an important infl uence in economic mobility of mi grant/minority groups. The two groups in this study are both citizens of the receiving society, and are entitled to the full range of benefits. This makes it possible to look at the influence of differences in benefits and provisions in each country. The receiving society plays a role in the opportunity structure of immigrants. To avoid too much complexity on this level, I selected two cities: Amsterdam in the Netherlands and New York City in the United States. The selection of cities is necessary to control for the labor market context of reception. The two cities have similarities in their immigration history, their economic development, and importantly, in contemporary changes in the structure of their labor markets. Both cities have experienced changes in occupational distributions with the shift from manufacturing to a service economy, and in both cities, the selected groups both form about ten percent of the total population. Due to these commonalities, I assume this part of the context of reception to be very similar for both groups. In the third context of reception, the ethnic communities, the strength of ethnic networks and ties is relevant for the incorporation of new migrants. Economic opportunities in ethnic entrepreneurship have stirred a discussion on the importance on this path of incorporation. As with the similarity of the labor market context of reception, the selected groups are similar in their ethnic context of reception. The first chapters discuss their colonization histories, their migration histories, their legal stars. the 0i of the group. the starting l Figure l-l ( TheOret From the from theta“, ll‘imt'in Cap, Elite-ts in the composition In ”'1 ChISH‘i status, the opportunities that they were able to create in the cities, and the composition. of the groups. I look at their economic position in the 19805 to give an impression of the starting point of mobility that is taken in this dissertation. Figure 1-1 Contexts of Reception .EtmicerpCIntext ofRecerxion Ldnor Maket Cbrtetd of Recenion \‘ . Econom'c Mobility .Walfae State Christel of Reception Defined as the Welfare State Context of Reception Theoretical Perspectives From the literature on immigration, from literature on economic mobility, and from literature on migration and mobility, several perspectives emerge. Aspects from Human Capital Theory are a recurring theme in all these literatures. I included these aspects in the composition of the group. As such, education levels, family composition, and time of arrival (or length of stay) are part of the analysis (Borjas 1994; Chiswick 1979; Kerckhof’f 1995; Treiman and Yip 1989). The n; hmamt Sweat ofthe strt utth the c elements in: grom 1945. W1 A thtr theories a Eager the LIE percer Huma the mairi l the mOtlt‘l Che«'acten 3110 n is p; The nature of the receiving economy and the changing labor market structure form a second string of discussion in these literatures. From World Systems Theory, Segmented Labor Market Theories, and Dual Labor Market Theories, the importance of the structure of the labor market emerges. With the changing global economy, and with the changed local economic systems of New York City and Amsterdam, elements from these theories have to be included. Unemployment levels and rates of job growth are the used indicators (Lipset and Bendix 1967; Sassen 2000; Wallerstein 1995; Wilson 1987). A third literature is more specific to immigrant labor market opportunities, and theories as the Ethnic Succession Model and the Ethnic Queue Model predict that the larger the percentage immigrants in a labor market, the higher the competition. Thus, the percentage of immigrants in each city is included (Ortiz 1991; Wilson 1996). Human Capital Theory and ‘Skills Mismatch’ theory both focus on education as the main road to occupational mobility. One major problem of including education in the model is that education is part of each context of reception: it is a group characteristic, it is linked to the labor market through the Skills Mismatch Theory, and it is part of the welfare state because of public funding of education, because of special provisions for minority groups in education in both countries, and because of specific job training programs adapted to needs of the labor market. The last set of measures is derived from theories on the welfare state. There is no real definition of ‘a’ welfare state, and I am using an ideal-typical classification derived from Esping-Andersen (1990). The main divisions between the three different types of welfare states are based on levels of de-commodification and degrees of c rcstratittcatttti pamcipation can dense inc strantieation 1 through malt itlerent ripe Heidenheimci l’ne prostsron plonsmns get stratification . PIOVlSlOILs 01‘ he less. This l 13W mailer leiel mm bt‘i measure en ti mmums on t '1‘ an exampt Researcl restratification. The first is a measure of dependence on income from labor market participation. The term de-commodification applies to the process by which people can derive income from sources outside of the labor market. The degree of stratification links to the ways the welfare state reshapes the income distribution through taxation and other provisions. In this framework, the United States is a different type of nation-state than the Netherlands (Esping-Anderson 1990; Flora and Heidenheimer 1981; Hamnet 1998; Kloosterman 1994; Visser and Hemerijck 1997). The provisions in the United States are far less than those in the Netherlands. Larger provisions generated by the welfare state will lead to more redistribution and less stratification within society, and to more equality. A side effect of these larger provisions of the welfare state is in that incentives to participate in the labor force can be less. This is turn can lead to a population that is excluded from participation in the labor market, this population might not feel the need or possibility to work in lower level jobs, because of the provisions they receive through the welfare state. Besides a measure on the amount of public spending (as percentage of GDP), I have included measures on taxation and on the amount of public or private spending in health. care (as an example of the separation between state and market provisions). Research Questions This research is a systematic comparison between two groups at two locations in order to analyze a third variable, the impact of the welfare state, on the occupational mobility of the two groups. Such a systematic (quantitative) analysis has not been done yet. There is research looking at two or more countries, looking at immigrant ‘adaptation’ in different aspects (political, economic, cultural). Most of this research is of a smaller has been rcsca BioemraadI. t ditTerent grou tire focus otIr Research t conclusions I secret}. it is c help economt had d:i.erertt welfare state Both groups 5 oeliare state more insight ; t in Canada 3 h dlifidtantanct hat oth er lImr hat, , . edliT‘Cren‘ dii“ . terenttal p, IRemain lest q “Ions refc is of a smaller scale, where a ‘same group’ in two countries was studied. Also, there has been research looking at multiple groups in one country (for example: Bloemraadl; (Hiebert, Collins and Spoonley 2003; Reitz 1998). Combining two different groups in two different countries over two different time periods is a more rare focus of research. Research on the role of the welfare state on migrant mobility has had conflicting conclusions. Because of the importance of economic mobility of new groups in society, it is crucial to understand the role of the welfare state: does it block or does it help economic mobility? Two groups moved to two different countries, where they had different access to welfare state benefits and provisions. Can these differences in welfare state provisions explain differences in economic mobility for both groups? Both groups studied are citizens of the country they moved to and have access to welfare state provisions in the same way as ‘native populations’. This might give more insight in a speculative statement made by (Hiebert 2002, p. 22): since refugees (in Canada) have a larger amount of settlement provisions, they are able to overcome disadvantages and they do not experience the long term effects of these disadvantages that other immigrant classes take longer to overcome. His study looks at groups that have differential access to welfare state provisions, my study examines the effects of differential provisions on a ‘similar group’. The main research question is answered through several sub-questions. The questions refer to the defined time period and to the specific location of each group: ' Paper presented at SSHA Annual Meeting 2002: "Government, Organizations and Immigrant Political Incorporation: A Tale of Two Countries." 1,; -Il Definh 1) Is the economic position of Surinamese better than the economic position of Puerto Ricans given their similar position in the 19805, or vice versa? 2) Has the economic position of Surinamese improved more than the economic position of Puerto Ricans or vice versa? 3) Which ‘context of reception’ has most impact on the economic position of each group in each city? 4) Is the welfare state (as the government context of reception) the most significant influence on their economic position? Definitions and Limitations I examine two groups that are not migrants in a legal sense, but that have migrated, as citizens, from one part of the official territory of the country to another. The label ‘migrant’ is used instead of ‘immigrant’ because of this reason. ‘Internal migrant’ (as opposed to international migrant or immigrant) is another label that could be applied, but internal migrant gives an implication of local movement while the groups had to cross multiple boundaries (as explained in chapters 2 and 3). Welfare state benefits and welfare state provisions are two different expressions that have been used interchangeably for the same set of arrangements. I use a very specific set of measures to test for the influence of the welfare state, and the descriptions of the welfare states are geared towards these measures. The measures reflect the institutional composition of each welfare state. International comparative research faces difficulties in term of definitions and measurements. A first obstacle lies in the ways groups are defined and counted. Both 10 groups migr: origin is th: with the deli aperson “a in the Count- identirs is n. 0“ an tr Cnited Stat; IlOl alto-3,5 ( tmustang 10 At‘IItcari anc pmpol'llOl‘l ( ”moor llJISIOnICan}, IIIICCSIFFI ( o treligiOUS, I (and 15) Iml Illt Nether]; ll'rigsn 199“, BecaUSe Puerto RICO m0 Very d“ Immnance groups migrated as citizens to the country of destination and the label ‘national origin’ is thus not a label that would select these groups. This is not the only issue with the delimitation of the group: a second way of counting is through asking where a person was born. With second (and third in the US) generations of the groups living in the countries of reception, this question is not adequate anymore. Commonly, identity is now a marker to define groups. On an international level, ‘identities’ creates another aspect of confusion. In the United States, identity is often linked to racial classification. These classifications do not always correspond with the identity of the person, nor do these classifications translate to other countries. The ‘one-drop-of-blood’ rule of the United States attaches African ancestry in a way that is not seen in many other countries. Thus, while a large proportion of Surinamese are of African ancestry, the consequence is not linked to a one-drop-of-blood rule and other ethnic markers are important. In the Netherlands, historically and contemporary, religion was the main societal divide, not ‘Afn'can ancestry’ (or other ‘racial’ definitions). In the past, Dutch society was divided into (religious) pillars, today, the Islam is a main label of ‘otherness’. While religion was (and is) important in the US, it did not (but might) have the consequences it had in the Netherlands (Rea, Wrench and Ouali 1999; T ubergen, Maas and Flap 2004; Wright 1990). Because of these problems, I have tried to include all people that migrated from Puerto Rico and Surinam. I do not select certain ‘parts’ of the groups, and as such, two very diverse groups are examined. This diversity makes the claim to the importance of structural aspects stronger. While discrimination cannot be ignored 11 tchapter 5 i. 100. A sew“ measured l COIiL’CIlOl‘t‘ most comlT : IistIactoT} Econon measured tl comparaiii internaitoii; mobility. e: Canzehtm Gal‘tzcboon not used co ihut not mt Pitture We (chapter 5), the groups’ heterogeneity suggests that other structural factors play a role too. A second difficulty in international comparative research is the way variables are measured. Educational categories differ, job descriptions are not the same, and data collections vary. These difficulties are addressed in chapter 4, and I have taken ‘the most common’ solutions. This does not mean that the solutions are completely satisfactory, or that they are safeguarded against critiques. Economic mobility is one of the variables that is measured using convention. It is measured through occupational mobility, a standard that is used most in international comparative social, economic, or occupational mobility. Over time and in international comparisons, there have been different definitions of occupational mobility, each with their benefits and constraints. For this dissertation, the use of occupational mobility was most feasible (Dessens et al. 2003; DiPrete et a1. 1997; Ganzeboom, Luijkx and Treiman 1989; Nakao 1992; Treiman 2001; Treiman and Ganzeboom 2000). This means that there is no monetary measure included: income is not used comparative difficulties on an international scale. Income might be reflected (but not measured) in the occupational scales, wealth does not show anywhere in the picture. Wealth, a form of financial capital, is thus not part of economic mobility explanations, an omission that might distort intergenerational mobility patterns. The welfare state has many measures that affect economic mobility. It is impossible to include all possible aspects, measures, or variables. The data sets have a limited number of respondents, which limits the number of variables that can be used. Secondly, not all data on the welfare state are available, and resources on a local level 12 are espeCIflllh not exhaustti- state can ha“ Structur The first i mention his Sunnam. and economic do Puerto Rico i migration his Chapter 1 Besides gent on migrant n Ch-5'CrimtIt‘.Etitt Opportunitie- on an intern; dimmlrtaiii O‘I‘hls dlSSet dISCIImInalll ChaPIL‘r “Nerds the lo the “filfa; ”999‘, are especially difficult to obtain. The measures presented in this dissertation are thus not exhaustive, but the results give an indication of the possible impact that a welfare state can have on the economic mobility of (immigrant) groups. Structure of the Dissertation The first chapters of this dissertation are descriptive analyses of the histories and migration histories of the two groups. Chapter two describes the colonial history of Surinam, and looks at the causes of increased migration in the mid-19705, a period of economic downturn. In chapter three I give an overview of the connections between Puerto Rico and the United States, after a period of Spanish colonization. The migration history of Puerto Ricans to New York City is the focus of the chapter. Chapter four is a theoretical chapter on immigration and economic mobility. Besides general migration and general economic mobility theories, specific theories on migrant mobility are included. Chapter five discusses segregation and discrimination. Discrimination is a societal aspect that affects the economic opportunities of both groups. Due to difficulties in the measurement of discrimination on an international comparative scale, I have not included a measure of discrimination in the model. That does not mean that discrimination is not an aspect of this dissertation. In this chapter I explain why a measure is not included and how discrimination and segregation would relate to the model. Chapter six is a background chapter on the welfare state. The chapter is geared towards the hypotheses and towards the model. The Dutch welfare state is contrasted to the welfare state in the United States, using the typology of Esping-Anderson (1990). 13 lnchapterse The city contexts methods are spc defined. These 1 results. and \\l1t’ are , esented in In chapter seven the literature and past findings on two groups are compared and the city contexts and ethnic group contexts are described. In chapter eight the methods are specified, the data sets are described, and the conceptual model is defined. These two chapters lead up to chapter nine, which includes the analyses and results, and where the general model is evaluated. The conclusions and the discussion are presented in chapter ten. 14 2. Surin The Sunna historical and ; mde, ‘ndence. Surinamese we ll; society Thi~ Sm’namese llll analyzed "Th8” Su The colonl the entry of the l650 to 16.671 Africa, The Br plantation‘colt racial hierareh tiulian militia abote the (We Wm mral force misfit/267‘ . ) [ R. , Chbsaet: RIO . 75m ‘ ma . hlip ‘6 “ 31¢an p'C'pulam‘ 2. Surinamese in the Netherlands The Surinamese position in the Netherlands and in Amsterdam is shaped by historical and contemporary events. The history of colonization, the changes towards independence, and global and local economic and social changes have impacted how Surinamese were received and how Surinamese were able to carve out their position in society. This chapter first describes the history of the Surinamese, and of Surinamese migration to the Netherlands. Second, their position in Amsterdam is analyzed. “The” Surinamese The colonial history of Surinam started in 1650 with English colonization. Before the entry of the Europeans, native populations were the Surinen and Arawakz. From 1650 to 1667 the British colonized the region and brought the first slaves over from Africa. The British rule ended when the Dutch took over control and established a plantation-colony based on the continued exploitation of slave labor. A social and racial hierarchy was quickly established: the Dutch planters were in control with a civilian militia, ‘colored’ people (kleurlingen) and free slaves were considered a step above the (West African) slaves. Some were able to escape Dutch control into the vast rural/forest areas and they are the forefathers of the Maroons (in Dutch: ‘Bosnegers’) (Amersfoort and Niekerk 2006; Niekerk 2000a; Oostindie 1988; Reubsaet, Kropman and Mulier 1982). ‘ 2 F rom: http://www.troon.org/suriname/suriname.html; http://www.sr.net/smet/InfoSurinarn/history.html. I did not find much information about the pre- colonial population in the literature. 15 With the end 1860 During the were recruited A med to ensure from Bnttsh-lndt freed-slates as la 33.000 lat anese Anot er group in 5 early as l630 Portuguese lnqui W . ' .ttllla. Oostindie 19%, Early emigre B)" the l93lls IellgtOUSlV. and c acolontal “at at Ilrttemattonal. C lll‘thh Native Su at this etent to t S . 1’81 Oter 700 3 \ “151% ‘ 1968 l With the end of slavery in sight, Chinese workers were contracted from 1853 till 1869. During the same time, Portuguese from Madeira and West-Indians (Barbados) were recruited. And when slavery ended in 1863, a different recruitment program started to ensure labor on the plantations. From 1873 till 1916 about 34,000 people from British-India, Hindus and Muslims, were brought to the colony to replace the freed-slaves as laborers. A third period of recruitment, from 1891 till 1931, brought 33,000 Javanese contract-workers from Netherlands-India to work on the plantations. Another group in Surinam are the Jews: there is evidence that Jews were in Surinam as early as 1639. The early Jewish groups came through Brazil (to escape the Portuguese Inquisition), later migrations came through the Netherlands (Niekerk 2000a; Oostindie 1988; Reubsaet, Kropman and Mulier 1982; Tjong Kim Sang 1995) Early emigration By the 19205 the population of Surinam hence was ethnically, racially, religiously, and socially very diverse. This diversity was exhibited 50 years before in a colonial way at the ‘Internationale, Koloniale en Uitvoerhandel- Tentoonstelling' (International, Colonial and Export Trade Exposition) in 1883: 28 Surinamese, among which Native Sunnamese3, Maroons, Creoles, and one Hindu, were literally displayed at this event, to show and educate the Dutch about Surinam. There had been Surinamese presence in the Netherlands before that period though: between 1729 and 1781 over 700 Surinamese migrated either by force to the Netherlands as servants, or 3 In 1957, the native population of Surinam was reduced to ‘weim'g talrijk ’: ‘little multitudinous’ (Amersfoort 1968). 16 for educational Muller 1933: S‘ [n the 63115 gmamese mls opportunll) ll" ot‘the elites m3 This migration Creole middle recruitment hat; Netherlands. tr little in 19571 Oostindie Witt The Dutch recruitment of here setere la h 1313 domestt abroad, to cou fulfill the dem .~ 3 at“ ~50 nur‘ got 'ernmenr pi for educational purposes (Amersfoort 1968; Niekerk 2000a; Reubsaet, KrOpman and Mulier 1982; Schuster 1999). In the early period of migration, roughly from the 19205 until the late 19605, most Surinamese migration to the Netherlands was for education: there was little opportunity for higher (post-elementary) education in Surinam, and talented children of the elites made the trip to the Netherlands (Bovenkerk 1975; Oostindie 1988)4. This migration consisted of children of Jewish Surinamese and of the upcoming Creole middle class. Some labor migration occurred in the 19305, and some labor recruitment happened in the 19505. In 1946 about 3,000 Surinamese lived in the Netherlands, in 1952 over a 100 Surinamese were registered at Dutch universities, while in 1957 this had increased to over 350 (Amersfoort 1968; Cottaar 2003; Oostindie 1988; Schuster 1999). The Dutch government did not favor labor recruitment from the colonies, the recruitment of nurses exemplifies the problematic position of the government. There were severe labor shortages in the Netherlands in the 19505, the search for nurses was first a domestic affair, but when the shortages remained, the search was extended abroad, to counties without colonial ties to the Netherlands. Only when this could not fulfill the demand, the government turned to Surinam in 1956. By 1958 there were about 250 nurses in training in the Netherlands from Surinam (Cottaar 2003). The government purposefully did not recruit from the colonies: they aimed for foreigners * 4 The non-fictional book ‘Sonny Boy’ by Annejet van der Zijl (Nijgh en van Ditmar, 2005) tells the beautifiJl and amazing story of one of these migrants, Waldemar Nods, who arrived in the Netherlands in 1927. He died towards the end of WWII on a German transportation — he was arrested and sent to a concentration camp in 1944 because he was active in the resistance. 17 [hat dld not l the contraett The legs the stereom Surinamese lS possrl‘let this exotten- The gmL fore but i can Ol‘l“ goo to s; rem p l Alttitthe behattor T that did not have the Dutch nationality so that they could be sent back at the end of the contracting period. The two recruitment programs were ended within a few years. The legal-nationality aspect was not the only reason not to recruit from Surinam: the stereotypes of Surinamese men and women were just as central in the choice. Surinamese women were considered in a ‘more positive’ way (if a positive stereotype is possible) than men: their difference was centered around exoticness. For the men, this exoticness translated into laziness and sexual aggressiveness: The Surinamese type, and with this I mean the Creole, which is a group of all kinds of ancestries, with the Negro as the most important forefather, is in general happy, uncomplicated, ofien childish naive, but also boastful and untruthful, but if he does not like somebody, he can be fanatically evil. He loves music and dancing and will take any opportunity to have a party, which will be very loud and involving a good amount of alcohol [. . . ]. In general, he does not work a lot, that is to say, he works slowly and he hates agricultural work, because it reminds him of the period of slavery of his ancestors (Schuster 1999, p. 125, my translation). Another part of the stereotype, specifically in the media, was on criminal behavior. This latter part became more pronounced as migration from Surinam increased. For the most part, Surinamese migration remained an education migration until economic and political circumstances in Surinam created a ‘push’ for mass migration. Increased emigation The Second World War was a period of economic growth for Surinam, but in the ten years after the war, a new kind of society developed: economic and ethnic-cultural interest groups competed for political power. This trend continued in the 19605 and the new political formation created a situation in which every election gave a shift in 18 the tthole hureaut‘ more strife het“ ee potter structures 1' groups that lost th At this time. It increasing. but the Migration to. the .\ letause of the eu Kropman and Mu then became the 1 although some Su percent of crimin; Netherlands 1 en': created an imprcs 39781979a. Scht the real crime rat Sunnam became acceptable “out, htfiterlands. Ant thscomposlllon ‘ Fear m‘I‘s’rated, “- mlmr ' the whole bureaucratic apparatus. This in turn elicited accusations of favoritism and more strife between the different interest groups. With the changes in government, power structures in the entire society changed, and migration became a trend for the groups that lost the elections. At this time, the Dutch economy was booming and labor shortages were increasing, but the government discouraged labor recruitment from Surinam. Migration to the Netherlands slowly increased through established connections and because of the economic and political problems (Bovenkerk 1975; Reubsaet, Kropman and Mulier 1982). In this period, the supposed criminality of Surinamese men became the focus of government and media discussions. Numbers reveal that although some Surinamese men (the estimates show that in 1965 between 4 and 5 percent of criminal court cases involved foreigners and pe0ple born outside the Netherlands!) engaged in criminal activity, the amount of attention that was given created an impression of severe criminality of Surinamese men (Bovenkerk 1978/ 1979a; Schuster 1999). Schuster (1999) gives a more plausible explanation than the real crime rates: in the 19605 discussions on restrictions on migration from Surinam became salient in government. One way to make these restrictions ‘more acceptable’ would be by criminalizing and problematizing Surinamese living in the Netherlands. And van Amersfoort (1968) describes that the crime numbers reflected the composition of the population: at that time, young men aged between 20 and 35 year migrated, women, children and the elderly were not as much part of the migration process yet. When controlling for these demographics, crime-rates of the 1.9 Surinamese In 1 population. Immigratio lion - as the 5 Netherlands. T twill. the "AC reported its ‘ea from other sou: tmnressron oft of die Commit hhile the Corr $1.. u tthe Oppogi their estimates Surinamese in the Netherlands were not (much) higher than those of the general population. Immigration numbers of that time indicate that there was not really an ‘enormous flow’ - as the government and the media claimed — of Surinamese moving to the Netherlands. The ‘Advies Commissie inzake migratie van Surinamers en Antillianen’ (1972), the ‘Advisory Committee on Migration by Surinamese and Antilleans’, reported its ‘careful estimates’, and assumed the numbers were probably higher, but from other sources it seems that the numbers were actually lowers. Table 2-1 gives an impression of the differences in numbers: in the first column, the ‘careful estimates’ of the Committee are reported, in the next two columns two other sources are cited. While the Committee thought its numbers were an underestimation, the data reveal that the opposite was the case. Even though the Committee included Antilleans in their estimates, this cannot account for the disparities". 5 The numbers from the Committee are higher than the numbers Reubseat et al. (1982) and Oostindie (I 98 8) produced. The data from Ruebsaet et al. is on the total of Surinamese in the Netherlands, Postindie reports net migration per year. The committee’s estimate of Antilleans in the Netherlands in 1972 was between 6,000 and 8,000: this number is too low to explain the differences 20 Table 2-1: Sur Adi Yea \etl . Sun l9tt4 W lit/36 . 1967 l968 l969 W \e—K \ Oserall. th in a specific tt records of in- ; Reubsam‘ KIO “ith the highs pgpillalion in l the [0‘3] POpul The migra SUIinameSe Sc IEUlllfiCauOn l mlgIEITOH bCI‘ Table 2-1: Surinamese migration to the Netherlands 1964-1972 Advisory Committee: Oostindie (1988): Ruebsaet (1982): Yearly migration to the Surinamese net migration Total of Surinamese in Netherlands of to the Netherlands the Netherlands Surinamese and Antilleans 1964 5,690 1965 7,000 yearly 1,825 11,000 1966 2,301 1967 2,425 1968 2,988 1969 9,448 4,370 1970 11,116 5,558 29,200 1972 12,813 6,313 37,000 55,000 (total 1972) 43,300 Overall, the variation in numbers exists because Surinamese were not registered in a specific way (since they are Dutch citizens), which also meant that there were no records of in- and out-migration (Amersfoort 1968; Oostindie 1988; Praag 2003; Reubsaet, Kropman and Mulier 1982; Schuster 1999; Tesser 1993)7. Regardless, even with the higher number that the Committee estimated, the Surinamese and Antillean population in the Netherlands was less than a half pro-mille (one on two thousand) of the total population (Schuster 1999). The migrants of this period were still mainly from the higher classes of Surinamese society. As with all migrations, chain-migration had started, and family reunification became more prominent. These processes determined most of the migration between 1968 and 1971. With the Dutch economy recovering from a small recession in 1966-1967, unskilled workers started to be drawn to the Netherlands. At the same time, elections in Surinam in 1967 changed the government and members of 7 This is a measuring problem throughout the years. There is no clear definition of how to label somebody as Surinamese. Some use ‘country of birth’, others use ‘country of birth of the head of household’, and sometimes ‘nationality’ is included in the statistics. Today, with more generations than the first generation, this problem has increased in complexity. 21 thhmmflhea immigrants heear illeahsaet. Kropn Although The treated. there is e imnngrantni \ltv triangle on the W Amsterdam. Den tecause of the ex tuners ln the ea immigrants gren 190913 "The rela immigrants caust mlé‘atlon after 1' “115 period \I ”ligation to the - lill-pollalttl}, the , ddueandedUC pan Oflhe SUl’ln; more Pimple bet. the Xflhcrlande ragl‘cllhci'k 197:. i d 51er pr e OCesg "pa-’1 0' 7 V t the Kits .4; \ the former state apparatus migrated. Overall, in this period, the selection of immigrants became more diverse, but the numbers were still relatively small (Reubsaet, Kropman and Mulier 1982). Although the numbers did not reflect the panic the government and the media created, there were some problems with the consequences of the increased immigration. Most Surinamese Dutch moved to areas in the Randstad, an imaginary triangle on the western side of the Netherlands that includes the cities Rotterdam, Amsterdam, Den Haag, and Utrecht. The most severe problem was in housing, because of the existing housing shortage and because of discrimination by house- owners. In the early 19705, while immigration was still low, resentment against the immigrants grew, and in 1972 the call for entry-restrictions increased (Schuster 1999)8: "The relatively small but steadily increasing stream of Surinamese immigrants caused the Dutch government to search for a means to call a halt to the migration after 1970" (Amersfoort and Penninx 1998, p. 48). This period was probably a crucial period in the consolidation of Surinamese migration to the Netherlands. The numbers of migrants increased, and more importantly, the diversity in background increased. Migration changed from a process of elite and education migration to a process of societal migration. As such, a larger part of the Surinamese society became knowledgeable of possibilities of migrating, more people became connected to emigrants that now were an important contact in the Netherlands, and more people had family that had made the cross over the ocean (Bovenkerk 1975). ‘_ 8 A similar process was and is taking place in regards to Antillean migration: the Dutch Antilles are still part of the Kingdom of the Netherlands. See Hulst, 2000. 22 As a 90“” aiailahle in S sistem. and I? This image “ i’rxnnanentl} and pom.) 9“ panern is tha‘ thlenjiiar EU according to . [for the Neth: information i that through 1 information r m} translatio impression th es‘ahlished t I Numbers melition of “Elation emu i,- , lmags of the new country As a colony of the Netherlands, information about the Netherlands was widely available in Surinam. The education system was based on the Dutch education system, and the colonial power was able to create a positive image of the Netherlands. This image was enhanced by the experiences of migrants. Often, people that returned (permanently or for vacation) were overwhelmed by the riches of the Netherlands, and portrayed a notion of a land of opportunity. As with other groups, a common pattern is that emigrants do not want to give an impression of failure (for example (Menjivar 2000). But even if they did give this image, they were not believed: according to one informant of Bovenkerk (1975), “whatever you say, they will leave [for the Netherlands] anyway” (p. 60, my translation). And another part of the information flow was distorted by other psychological processes: “it is certainly so that through processes of selection and reinterpretation for many emigrants the information portrayed is more rosy-colored than the reality” (Bovenkerk 1975, p. 64, my translation). In general, the information about the Netherlands gave a positive impression that motivated people to follow their ties and networks that were established (Bovenkerk 1975; Reubsaet, Kropman and Mulier 1982). The role of the state Numbers of immigrants are ofien not the main reason to restrict migration, the perception of a ‘threat’ can create a situation where the call for restrictions of migration emerges (Massey 1999). As the role of the govemment/state is already limited to control immigration in general, the role of the state is even more limited 23 iihen the mi Jopple Mt In 1970 article abou in 1954: "Ai cannot be c; h) lets" (Sc “as not as [ centered art drillilcan m “hat Ilme‘ S became indi when the migrating groups are citizens of the country they move to (Janoski 1998; Joppke 1998c). In 1970 the government tried to amend the Dutch Constitution to include an article about rights of citizenship (article 6, paragraph 2). The amendment was written in 1954: “All. Dutch are admitted into the Netherlands; they cannot be expelled. They cannot be cannot be obstructed to leave the country than in cases that are determined by law” (Schuster 1999, p. 135, my translation). In the early 19705 this formulation was not as popular as when it had been written, and the discussions in parliament centered around how to alter the amendment so that restrictions on Surinamese and Antillean migration could be constitutional. These discussions continued till 1976, by that time, Surinamese migration would not be affected by this change since Surinam became independent from the Netherlands in 19759 (Amersfoort and Penninx 1998). Independence After the 1973 elections the NPK'O-govemment suddenly started to negotiate with The Netherlands about the independence of Suriname. The negotiations succeeded. Suriname became independent at November 25, 1975 and the Netherlands agreed to support the young country for ten years with a total amount of four billion guilders (Tjong Kim Sang 1995, online). There had not been many references to independence before or during the elections, and according to Schuster (1999) the announcement came as a surprise for the Surinamese. The Dutch government also had declared, as part of coalition 9 Migration from the Dutch Antilles could potentially be affected by this change in the amendment. The right to entry is not explicitly linked to citizenship and “The lawyer Willems has concluded from the proceedings that government and parliament have kept a ‘clean option’ to retain the possibility to exclude Antilleans, but to guarantee the entry of Dutch Dutch” (Schuster 1999, p. 138, my translation) ‘0 Nationale Partij Kombr'narr'e: National Party Combination. 24 formation in l' fen aspects th; and during 19". important. and Netiiorls in tl become cheap laters of Socic ethnic basis of Surinamese m had migrated r “here ”1636 l a JmancSe-Sum Pmmaribo ltl Mal gmUPS s Kropman and In the NCI more ficrc 61}, the consmum malignant." re: “‘38 an idea] ‘ Bill as oil 1994'): m‘s’Tat for differeht r formation in 1972 that Surinam had to become independent before 1976. There are a few aspects that might explain the Dutch desire for an independent Surinam. In 1972 and during 1973 autonomous (auto-dynamic) forces of migration became more important, and even though the Dutch economy was shrinking, migration increased. Networks in the Netherlands made it easier to make the crossing, and airfares had become cheaper. With the worsening economic situation in Surinam, people from all layers of society migrated. And the elections of 1973 gave another political turn: the ethnic basis of politics became even more important and now many Hindustani- Surinamese migrated. In earlier periods, J ewish-Surinamese and Creole-Surinamese had migrated to a larger extent: the earlier migrations were mainly from the cities, where these latter groups were overrepresented. Some Hindustani-Surinamese and J avanese—Surinamese left in the earlier periods, mainly by migrating first to Paramaribo (the capital of Surinam) and then to the Netherlands. In this period, the rural groups started to leave directly for the Netherlands (Amersfoort 1968; Reubsaet, Kropman and Mulier 1982). In the Netherlands, the opposition to low numbers of migrants now expanded to a more fiercely opposition to increased migration. The parliamentary discussions about the constitutional amendment on the automatic entry of people with the Dutch nationality reflected this opposition, and for the Dutch government, independence was an ideal way to curtail migration from Surinam. But as often, there were unintended consequences of legal measures (Calavita 1994).: migration from Surinam increased dramatically in 1974 and 1975. People lefi for different reasons but two stand out: the economic and political uncertainty of the 25 afierrnath o Surinamese becoming 5 before it its tReubsaet. l Fore Suri fron‘ Shot font the} The year miligation. tl Education, E anew migm that yearE 3U KIOpman an Afler 19; imml‘s’Tantg t iiben the Sui inflation, Afi Sll'il'lgent_ BEE llllt'lOCumcmc mi»a m no“ Stat that Ofthe ear aftermath of independence, and the change in nationality afier independence. While Surinamese in the Netherlands had the choice between staying Dutch nationals and becoming Surinamese, Surinamese in Surinam did not have that option. People left before it was announced that there would be a transition period from 1975 till 1980 (Reubsaet, Kropman and Mulier 1982): Foreigners with a Dutch passport: such was the status of the Surinamese at the time that the mass migration to the Netherlands from the former colony of Surinam [or Dutch Guyana] took place. Shortly before Surinam’s independence in 1975, many left for the former mother country with high hopes. Upon arrival, however, they were shocked to discover that the Netherlands was not the paradise they had always believed it to be (N iekerk 2000b, p. 64). The years afier independence showed optimism and a sharp decrease in migration: the main motives of migration were now family reunification and education. But quickly emerging economic and political problems in Surinam created a new migration flow started in 1978: 1980 was dooming and after September 1St of that year, Surinamese entering the Netherlands would need a visa (Reubsaet, Kropman and Mulier 1982). After 1980 migration became a more steady process, with about 4,000 to 6,000 immigrants to the Netherlands a year. There was a migration peak in 1993 (9000) when the Surinamese economy experienced another problematic period with high inflation. After 1993, migration again decreased, Dutch border controls became more stringent. Besides reducing migration to around 3000 a year, this also increased undocumented migration from Surinam to the Netherlands (Niekerk 2000a). "Official migration statistics show that the legal migration in the early 19905 is not lower than that of the early 19705” (Amersfoort and Penninx 1998, p. 51). Table 2-2 gives an 26 OWTHL’“ Since 197 Table 2-2 Netherla T. overview of the migration rates and the Surinamese population in the Netherlands since 1972: Table 2-2: Net migration rates and the total number of Surinamese in the Netherlands 1973-2003. Year Net Total of Surinamese Migrition in the Netherlands 1973 9,035 52,500 1974 15,674 68,500 1975 36,537 105,400 1976 612 106,500 1977 1,368 108,500 1978 4,710 113,000 1979 15,789 1980 16,705 **"‘ 146,000 1985 3,768 ***181,000 From: Oostindie From: Reubsaet (1982) Q988) 1990 *237,000 1995 1525 **276,000 2000 2808 2003 2501 ”321,000 From: CBS Statline * From: Tesser, 1993; "”" From of birth lllggal'fl : Praag, 2003; ***From: Vermeulen and Penninx, 2000, by country The numbers of illegal Surinamese migrants do not seem large. Research in Rotterdam gave nine illegal Surinamese out of 169 respondents - which would give under 600 illegal Surinamese in Rotterdam. In Amsterdam, if these numbers translate from city to city, the amount would be larger, but still less than a thousand people (Burgers 1999; Engbersen 1999; Niekerk 2000a): The population that is already settled in the Netherlands creates a network for family forming and family-reunification migration. After 1982 there were applications for asylum from Surinam [. . .] A fairly large number of Surinamese do not return from a tourist visit to the 27 Neil estii M’cra‘; ‘ #— The fit: Arnsterdarr the port of 19681. 1n tl increased population Percentage Ps‘Ople. an. lAmCTSi‘OC APosters.“ lAmCl’Sfm 0f the mClUded . Were 00m Cl‘i‘n 111011 llle major fncnds 0r Netherlands and become illegal immigrant, although it is impossible to estimate the amount of illegal stays (Heelsum 1997, p. 6-7)1 . Mgration to Amsterdam The first Surinamese immigrants, of the Creole elite, went mainly to Leiden and Amsterdam for study. Another group of early immigrants were sailors who arrived in the port of Amsterdam. In 1966, 6,496 Surinamese lived in Amsterdam12 (Amersfoort 1968). In the 19705 these numbers increased as migration to the Netherlands increased: while in 1973 the Surinamese population formed only 2.1 percent of the population of Amsterdam, this increased to 3.7 percent in 1982, and in 1988 the percentage reached 7.2 percent. In 1991 the total population of the city was 702,731 people, and of these, 58,010 were of Surinamese descent, a little over 12 percent (Amersfoort 1992)l3 . In 1999 25 percent of all Surinamese in the Netherlands lived in Amsterdam, 65,000 people, and of them, about 70 percent are of Creole descent (Amersfoort and Cortie 1996; Niekerk 2000a; Praag 2003). Of the migrants that arrived before 1963, 85 percent were from Paramaribo. This included ‘step—migrants’ that had moved from rural Surinam to the capitol. Creoles were dominant, and they were employees with lower educational levels. As such, even though the image of elite migration in the earlier migration periods was given, the majority was not of the highest parts of society. At arrival, most stayed with friends or family, mainly in Amsterdam. Between 1963 and 1967 there was a slight " In 1980 Desi Bouterse lead a military coup, on March 11 1982 a counter-coup failed. Since protests and calls for democratization continued, the military arrested 16 opponents of their regime on the night of 8 December, and executed 15. '2 For comparison, the other Dutch main cities had much lower numbers of Surinamese living there: Rotterdam: 2,002; Den Haag: 1,294; Utrecht 1,231. ‘3 1n the case of this article: Surinamese citizens, Dutch born in Surinam + children from head of family born in Surinam. 28 increase in concentrate came from areas. and t Creole dest became loi Hindustani trend conti more and g 1976 and t lReubSaet. The p increase in migrants with lower vocational education, and migration was less concentrated to Amsterdam. From 1968 until 1971 , the majority of migrants still came from Paramaribo, but there was an increase of people from the surrounding areas, and there was an increase in Hindustani migration, half of the migrants was of Creole descent in this period. The proportion of men increased, and educational levels became lower. Between 1972 and 1973 almost 40 percent of the migrants were Hindustani, education levels decreased again, while unemployment increased. This trend continued between 1974 and 1975 but with slightly more women migrating more and a minor increase in the percentage of rural migrants. Similarly, between 1976 and 1978, the differences with the preceding migration period were minimal (Reubsaet, Kropman and Mulier 1982). The position of the Surinamese in 1982 The Dutch welfare state has strongly mitigated some of the most insidious social effects of deindustrialization, particularly in the housing market (Niekerk 2004, p. 162). The position of Surinamese in the Netherlands in the late 19705 and early 19805 has been described in a series of reports (Reubsaet 1982; Reubsaet and Kropman 1983; Reubsaet, Kropman and Mulier 1982). These reports give a meticulous description of all aspects of Surinamese migration and integration. Unfortunately, the data are not comparable with data from later surveys as the SPVAM (Dagevos 1998). They are a good source of information to describe the situation in the decade after the peak of migration. M Sociale positie en voorzieningengebmik allochlonen: Social Position and Use of Provisions by Allochtones. The survey is explained in chapter 8. 29 1n 1‘ and as a populati} tfor the g fourth lti L'trechti 19831. Sectt In 1982 there were a little more Surinamese men than women in the Netherlands, and as a group, the population was much younger than the ‘white’ Dutch population’s. Of all the Surinamese men and women, 79 percent was younger than 35 (for the general population, this percentage was a little over 55 percent). Over three- fourth lived in the three most western provinces (North Holland, South Holland, and Utrecht) and 60 percent lived in the four big cities (Reubsaet, Kropman and Mulier 1982) The ethnic background of all Surinamese in the Netherlands is shown in Table 2-3. Of the Creole, 49 percent lived in Amsterdam and 52 percent in Rotterdam. Numbers for the Hindustani are not given16 (Reubsaet, Kropman and Mulier 1982). Table 2-3: Ethnic background of Surinamese in the Netherlands, 198217 Group Percentage Creole 39 Hindustani 40 Javanese Surinamese Chinese Surinamese Native Surinamese Maroon (Boslana' Cream!) flNLhL/t Labor Market Position The situation of migrants on the labor market is of essential meaning for their existence in a new society. This situation, in large part, determines their social-economic position. And this social-economic position, in turn, determines to a strong degree their position in other sectors of society, as housing and education (Reubsaet, Kropman and Mulier 1982, p. 115, my translation). ‘5 The numbers are contrasted to the white Dutch population, excluding guest workers but including Indonesian-Dutch: Indonesian-Dutch are not seen as a minority. ‘6 Which in-itself could be seen as an indication of a ‘bias’ in the Netherlands towards Creole Surinamese: even thought the percentage of the population was lower than that of the Hindustani, and even though their socio-economic position was better, was more attention given to the Creole Surinamese because this is a ‘black population"? ‘7 The percentage of Jewish-Surinamese is not available 30 In 1983 unemplo}e unemplosn educated S Among the were Aft 111‘th 0pm: ‘stable C01“ amrnore 1 the labor 1‘. Working u Oftl’te 10b that 4j skilled m, emf/‘10} ed WCcntag Table 2-4 K r? In 1982 almost 25 percent of the Surinamese in the Netherlands were unemployed, for the Dutch population this was 5.5 percent. In Amsterdam, unemployment was slightly higher at 26 percent, and especially young, lower educated Surinamese that came to the Netherlands after 1974 were unemployed. Among the lower educated, with elementary school or less, unemployment was severe. After 1972 general unemployment rose, and migrants with less education had little opportunity on the labor market. Of the employed, 37 percent was considered a ‘stable core’ (in the survey) that did not experience problems with employment anymore. Nineteen percent of the employed formed a risk group, whose position in the labor market was not secure, and 19 percent was in a bad employment position, working under their education level in jobs that did not provide much security. Of the total employed, 39 percent was what was considered a ‘lower employee’ a job that 41 percent also held in Surinam. Twenty-eight percent was employed as skilled manual worker, 20 percent as unskilled manual worker. Only 11 percent was employed in the higher regions of the labor market, a number that reflects the percentage of people working in these parts of the labor market in Surinam (12 percent): Table 2-4: Labor market participation working population, in percentages. Surinamese Sufinamese Dutch men Dutch men women women Lower positions 32 53 16 45 Skilled manual 32 19 28 1 l labor Unskilled 21 18 6 5 manual labor Middle and 11 11 38 32 higher positions From: Ruebsaet, Kropman, Mulier, 1982. 31 ln gener: strata of the lcxels and 1; groups In St the lQSiIs n concentratc semce lhc. for older p WWW (C al. 200]. p Roelandi g. dormsizin E11 In general, in the early 19805, many Surinamese migrants worked in the lower strata of the Dutch labor market. Several reasons were given: the lower education levels and lack of experience in the (Dutch) labor market. Competition with other groups in society and discrimination are additional factors. Overall, the situation in the 19805 resembled a dual labor market, where Surinamese employment was concentrated in the secondary labor segment. People mainly worked in the industrial, service (health care), and trade sectors of society, where the situation was more stable for older people. About half of the people surveyed had a stable job, regardless of job-type (Choenni and Cain 1995; Dagevos 2001; Gras and Bovenkerk 1999; Raes et al. 2002; Reubsaet and Kropman 1983; Reubsaet, Kropman and Mulier 1982; Roelandt and Veenman 1994; Venema 1992). The restructuring and economic downsizing of the 19805 affected minorities to a greater extent than other groups: Ethnic minorities suffered more than any other group from consequent economic restructuring and changed labour market conditions. In the 19805 large number of unskilled and poorly educated workers, minority workers as well as indigenous ones, lost their jobs and joined the ranks of the long-term unemployed. Their position in the labour market has changed from that of being at the bottom of the job structure to that of being excluded from the labour market all together (Roelandt and Veenman 1994, p. 48). Discrimination is another recurrent reason for the high unemployment percentages of Surinamese and other minorities. The Netherlands had only recently (1979) acknowledged that it might be a ‘country of immigration’ instead of a country of emigration, and minority policies or other measures to counter discrimination or selective hiring were not implemented (yet). For example, when employers sought to 32 mnempk 'g sf, ‘(rt’i qu1ckl}. bl indigenouj Reubsaet I In the : erupt"; .{1 my. Netherlant CXpencht and of tho: discnmina employed Forlor un€01pl0._\r Again, 01% indicated r 1983.1. The hié COmpared I an Obstacle cmain Rim 2: At the . Ger hire employees through employment agencies“, they were able to add the criterion of ‘g.s.’. ‘Geen Surinamers ’: ‘no Surinamese’. The written practice was outlawed quickly, but “about 30% of Dutch employers in the private sector prefer to recruit indigenous Dutch employees” (Roelandt and Veenman 1994, p. 48; Dagevos 1998; Reubsaet 1982). In the surveys of Ruebsaet et al., ‘experienced discrimination while looking for employment’ was one of the questions. Of those with a job who arrived in the Netherlands before 1963, 10 percent of respondents indicated that they had experienced discrimination while looking for a job. This number changed drastically, and of those that arrived between 1976 and 1978, 31 percent indicated that discrimination was part of their job-search. Overall, in 1983, 24 percent of the employed responded that some form of discrimination was an aspect. For long-term unemployed, lack of education was the biggest determinant of their unemployment, combined with a lack of information of the Dutch labor market. Again, discrimination is a (partial) reason for their unemployment, 41 percent indicated negative treatment by potential employers (Reubsaet, Kropman and Mulier 1982) The high levels of unemployment for Surinamese in Amsterdam can be attributed to similar causes as their labor market position. Education levels were lower compared to Dutch education levels, and with the economic restructuring this became an obstacle. Before restructuring, Surinamese (and other minorities) were hired for certain kinds of jobs. With automation, economic depression, and the increasing ‘8 At the ‘Gemeentelijk ArbeidsBureau’: the ‘municipal job bank’. 33 transfer of] increased 5 F1113”). Sui ofeconomr qualrfieatm the change: {men the o Oxerall FIOblCmgm. leaTS and 0 While; lmr fun the Sur transfer of production, their unemployment became endemic. Job competition increased: since more ‘whites’ were unemployed, they took more kinds of jobs. Finally, Surinamese (and other immigrants) did not profit from the subsequent period of economic growth: the period before 1973 had seen an increase in educational qualifications, and in service sector employment. With the economic downturn and the changes in manufacturing, migrants lost their jobs while they had not been able or given the opportunity to benefit from these changes (Choenni and Cain 1995). Overall, of Surinamese migrants in the Netherlands that were in a non- problematic position, 60 percent had migrated before 1968, were people aged 50 years and older, and had higher education levels (Roelandt and Veenman 1994), while: Immigrants with little education and few occupational skills, who formed the majority of the new arrivals, have found few openings in the labour market. Lower-class Creole and Hindustani-speaking Surinamese, [as well as the worst educated Antilleans and Arubans], have produced a second generation that may encounter social rejection because these ethnic groups as a whole have been typified as unfit for 'legitimate' work (Gras and Bovenkerk 1999, p.95). Poverty In the Netherlands job loss and poverty do not have the same meanings as in the United States: poverty is ‘social poverty’, where the financial consequences are not as severe, but the social consequences of exclusion and non-participation are similar. “This situation is not referring to a financially disadvantaged category with incomes below some absolute and objective poverty line, but to a category without social and economic skills” (Roelandt and Veenman 1994, p. 49). 34 Sunn. educated After 197- see future *herr educ. Netherlanr posmon The Dr m the L'rm cOnllnuous (careful) Ol education ( Education levels Surinamese that came to the Netherlands before the 19705 were generally higher educated: one of the migration motives was to get more education in the Netherlands. After 1974, there is a trend in lower educated migrants. In 1983 it was still difficult to see future educational prospects: the first generation migrants mainly had finished their education in Surinam, their children (either born in Surinam or in the Netherlands) were still too young to generate conclusions about their educational position. The Dutch education system was fairly complex. There was no middle school as in the United States, and after elementary school there were several options for continuous education. For elementary school aged children (6 to 12), the only (careful) observation was that most Surinamese children were enrolled in regular education (and a relatively low percentage in special education). For children of (Dutch) high school. age (12-18) Surinamese children were overrepresented in vocational training and underrepresented in middle and higher educational tracks. For those Surinamese children born in the Netherlands the differences with the ‘native Dutch’ population were smaller. The age difference in high schools could be explained by the difficulty of starting an education in Surinam and continuing that education in the Netherlands (22 percent of the respondents had difficulty with the connection between the education systems). The selection criteria were relatively strict, student were forced to ‘double’ a year (18 percent), and although the system in Surinam was based on the Dutch educational system, there were problems with the pace of education (11 percent), more difficult 35 subjects 130 discriminant of the respor Surinam: and boys in ' from the rur mi [3311011 t}: Table 2-5: I percentage —\ .\ Less [l films £6135 Ma LO“ 6] 49939 \lrdd‘ 1393; 5 ngh: subjects (20 percent), and differences in style of teaching (12 percent). Finally, discrimination was experienced by 20 percent and language problems by 23 percent of the respondents. 9 Surinamese girls generally continued their education in ‘middle higher education and boys in vocational training. Students with the most problems are those who came from the rural areas of Surinam and had lived in Paramaribo 5 to 10 years before migration (Reubsaet 1982; Reubsaet, Kropman and Mulier 1982): Table 2-5: Educational level at point of migration, per migration period (in percentages), migrants 13 years of age and older19 Before 1963- 1968- 1972- 1974- 1976- Total 1963 1967 1971 1973 1975 1978 Less than 11 7 9 l 1 19 17 14 elementary Elementary” 17 37 49 63 54 50 50 mulo 44 25 21 8 12 13 16 Lower 9 13 9 6 6 5 7 vocational Middle 13 12 10 10 7 10 9 (vocational) Higher 4 5 2 2 2 3 2 education unknown 2 2 Employment is related to education levels: of those with less than elementary school, 35 percent is employed. A small majority of Surinamese with ‘elementary school completed’ has work: 56 percent. And dependent on migration period, people that migrated before 1974, with lower vocational training completed, are employed K 19 The categories are not the same as the categories used in chapter 8, here I display as much izgformation, without collapsing the categories. Glo-a, glo-b, ulo II, ulo HI combined (‘regular lower education’ a and b, ‘extended lower education’ II and III). I have separated the number for ‘mulo’ (more extended lower education) in the table, because it is more comparable to high school than to elementary school in the US. Appendix 1 shows the Surinamese educational system. 36 fbrStipe less emp the empr unempro cities M ‘adequa’. all popuj The ilk? firgt (class; h Wan the neighbor “em up. Ofihe 10 large 3U] mama” acquirec for 80 percent. Those with the same education level, but that migrated after 1974, are less employed: below 68 percent. In general, the later the migration period, the lower the employment percentages. In the higher education population, there is little unemployment among Surinamese migrants. Housing Most Surinamese migrated to the western parts of the Netherlands, to the big cities where housing problems were notorious. Housing shortages and lack of ‘adequate’21 housing was a problem of the cities, and specifically of Amsterdam, for all populations. The Housing Act of 1901 implemented regulations and minimum standards for the first time: before, building was hastily and off poor quality. The Act gave rise to (class) homogeneous neighborhoods in the early 20th century. After the Second World War, the city controlled shortages in housing by freezing rents in the old neighborhoods. This had as a consequence that people stayed even when their wages went up, and that the construction of new housing was slow and expensive. Because of the low rents, landlords were not inclined to renovate, and over time this gave a large supply of poor maintained, cheap housing on the one hand, and expensive, well- maintained new constructions. In the old neighborhoods, “these neighbourhoods acquired a very stable character, with slowly degrading buildings and a gradually aging population” (Klerk and Amersfoort 1988, p. 152). When the old populations 2‘ Allocation rules only apply for houses that are short in supply: they do not apply for houses below certain standard (old city), nor above a certain rent level (market). The distribution of the houses (that do fall under the rules) is by family size and wage level. These characteristics give urgency for placement, those that are eligible are placed on a waiting list. 37 (finally the )0Ut‘lé was little housrng t (Amerst‘c The r 0f the )‘Oi position this syste' offered a housmg I} Anoti With Olhe tended to and Amer neighborh 100 far fro. 1983; ReL The B gotta] hou oflts Dun reached m and 10.004; (‘finally’) left, the change of the social composition of the neighborhoods was drastic, the young, the poor, and immigrants moved in. Private investors moved out, there was little money for investments in the first place because of the low rents, and the housing became municipal. Nearly all middle-class families moved out of Amsterdam (Amersfoort 1992; Klerk and Amersfoort 1988; Valkonet-Freeman 1978/ 1979). The position of the Surinamese (and other immigrations) was aggravated because of the younger age of the population and the generally weaker socio-economic position. Their recent arrival affected their position on ‘waiting lists’ for housing. In this system, the worse housing is easier available, people move as soon as they are offered a better place. Thus, ‘late-comers’ on the housing market have to accept housing that might not fulfill their needs. Another aspect in the housing situation was the desire to live in neighborhoods with other Surinamese: for social contacts and help. “Upon their arrival immigrants tended to settle near relatives and friends and often shared a house with them” (Klerk and Amersfoort 1988, p. 147). Guest workers settled in the 19th century (laborer) neighborhoods around the center, Surinamese also concentrated in the Bijlmer, not too far from the city-center (Klerk and Amersfoort 1988; Reubsaet and Kropman 1983; Reubsaet, Kropman and Mulier 1982; Valkonet-Freeman 1978/ 1979). The Bijlmermeer was constructed by the City of Amsterdam, as the latest of many social housing developments. It was a series of hi gh-rises that composed almost a city of its own. The area was still under construction when Surinamese immigration reached its peak, but yet, by 1983, the Bijlmermeer housed 53,000 thousand people, and 10,000 of the 35,000 Surinamese that lived in the limits of Amsterdam. In 1982, 38 seven percent of the population of the city lived in the Bijlmermeer, but 28 percent of the Surinamese p0pulation of the city. The percentage of Surinamese living in 19th century residential areas was also higher than for the rest of the population, but it decreased over time. A final area with a higher concentration of Surinamese was in ‘central location built between 1906 and 1930’. The main shift with increased migration was from the center, from lodging houses and furnished rooms, and from cheap and low-quality housing built before 1901 (Amersfoort 1992; Klerk and Amersfoort 1988): The Surinamese have succeeded better than other migrants in avoiding the poorest housing. Initially they appear to follow the general immigrant pattern of entering the housing market at the end of the queue. Nevertheless within a short time, they have cast their nets far more widely and have modestly but unmistakably got their catch (Klerk and Amersfoort 1988, p. 159). The large share of public housing in Amsterdam with the strong bureaucratic and political control over the allocation process has resulted in an ambiguous picture: on the one hand, the indices of dissimilarity for the city and for the Surinamese population are not as high as the indices for cities in the US. When the Bijlmermeer is excluded, the dispersion of the group in the city is even. On the other hand, with the inclusion of the Bijlmermeer in the computation of the index of dissimilarity for Surinamese in Amsterdam, the numbers are as shown in Table 2-6 (Klerk and Amersfoort 1988). The demographic structure of the Surinamese population explains their settlement pattern partially, the city is segregated on life-style and age, which is also reflected in the patterns of Surinamese housing. That does not mean that the housing market is a fair institution. Officially, discrimination and refusal of certain renters does not happen, but unofficially neighborhoods try to avoid non-white 39 inhabitants. ' finall}. insti in the cit}. I exclusion of Klerk and A 0\erall. small Rent? eSpfi‘mll} n the aierage had nor im; AmSlEl’danj Reubsaet 2 Table 2-6: h u. .._ inhabitants. The municipal allocation process can have biases towards minorities. Finally, institutional practices work out discriminatory towards groups that arrive new in the city. The waiting lists for certain areas and certain properties can result in exclusion of new groups in town (Bolle, Dijk and Hetebij 1978/1979; Essed 1991; Klerk and Amersfoort 1988; Valkonet-Freeman 1978/ 1979). Overall, the Surinamese often lived in high-rises or apartments that were too small. Rents were high, while incomes were low, and fellow-countrymen were, especially in the early stages of the migration, accommodated. Rents were higher than the average, but subsidies were allocated. In the ten years of migration, this situation had not improved significantly, but that also reflects the general housing situation in Amsterdam at the time. Housing issues were especially difficult for single men. (Reubsaet and Kropman 1983). Table 2-6: Index of dissimilarity for Surinamese in Amsterdam 1973-1981 Year Index of Dissimilarity 1973 29.5 1974 31.4 1975 32.7 1976 31.2 1977 29.2 1978 27.7 1979 27.0 1980 27.6 1981 288 From: Klerk and Amersfoort, 1988, p. 160. Family composition ‘Single parent families’ have a higher risk of living in poverty. Not only is there often only one breadwinner, the breadwinner is often female, and women have lower 40 earnings t both in Cr The re Suriname- headednei terminolo these “or less mone 0PPOrtuni in the lite Chances 0 Smith 195 \"ener SUtiname my trans]; are instiiu farmll-tie the insiab adapting, cOntempo eapnah-Sm EtonOmiC earnings than men. Single parent families exist more frequent among Surinamese, both in Creole and Hindu groups. The reasons for the larger percentage of female-headed households among Surinamese in Amsterdam are multiple. In the discussion around poverty and female headedness, there is often a stigma or ideological aspect attached (for example, terminology as ‘welfare queens’ or more subtle references to the lack of morals of these women). Unfortunately, the core of the discussion should be that woman make less money, and thus, that by being by themselves, they will have less economic opportunities than single men (with or without children). The predominant discussion in the literature on female-headed households in the United States is on the increased chances of poverty (Barry Figueroa 1991; Pérez y Gonzalez 2000; Rivera-Batiz 1991; Smith 1988). Venema (1992) uses the frequency of female-headed households among Surinamese without a “Eurocentric, male-centric, and middle class standpoint” (p. 28, my translation) through historical and synchronic interpretations: in Surinam, there are institutionalized alternatives to marriage. Household compositions are flexible and family-ties are strong. The role of the ‘inheritance from the past’ through slavery and the instability of families in slavery is not clear: are female-headed households an adaptation to circumstances or is there another ‘causal relation”? Important are the contemporary economic position of men and their ‘role’ of breadwinner: “peripheral capitalism, wage labor, migratory labor, the unequal sex-ratio, unemployment, and economic uncertainty” (p. 32, my translation) create a situation where matrifocality does not necessary translate into a problematic family situation. The father mi ght not 41 bEllHnt: family. if Son In 19 become ; minorir} Nfi‘lherla] categorie bi Emmi), coloma] l and c, pn COlOma’ I fr0m the f ‘\'it’W'lt-t,,, industrial. Penninx 3 be living in the household, but quite often that does not mean he is absent from the family (Sansone 1992; Venema 1992). Some Post-1983 Remarks Nevertheless, despite all of these public and private efforts, the minorities' disadvantage in the labour market is increasing rather than diminishing. Up to now, the employment policies have been less than successful; positive action programmes do not seem to result in companies hiring more workers with ethnic backgrounds and 60,000 jobs for minorities, promised in the STAR agreement, will not be realised on time (Gras and Bovenkerk 1999, p. 98). In 1983, the Dutch government officially acknowledged that the Netherlands had become a country of immigration instead of an emigration country. With that, a minority policy was deve10ped, aimed at improving the position of minorities in the Netherlands, while ‘tolerating’ the multicultural society. There are three main categories of immigrants in the Netherlands: a) foreigners from industrialized nations b) groups that entered the Netherlands through guest worker contracts or through the colonial history 0) others. Minority policy focused around groups from categories b and c, provided the group is in a weak socioeconomic position. As such, while colonial migrants are part of category b, Chinese and Indonesian Dutch are excluded from the Minority Policy. Additionally, the city of Amsterdam has a Nieuwkomersbeleid, ‘Policy for New Inhabitants’ aimed at all people from non- industrialized countries (Choenni 1997; Gras and Bovenkerk 1999; Heelsum 1997; Penninx 2005; Vermeulen and Penninx 1994). Part of the measures that were implemented to improve the position of minorities on the labor market were the EMO-plans, to promote employment of ethnic minorities in the government: 5 percent of the government employment is allocated to 42 mlIIOIIIIr 1993. Hr emplme 1035 an Mini “3pm er there “a ENUUDQ EQUIR Ii Out in 1g minoritn Said to h needed i also m C mecOnC diSCUSSH Dutch ‘c back'e’TOi much an \ MW minorities (1995), of these jobs, more than half were filled by Surinamese (Dagevos 1998; Heelsum 1997). The STAR-agreement of 1990, between government and employers, was intended to create positive action for minorities in the private sector (Gras and Bovenkerk 1999; Vermeulen and Penninx 1994). Minority policy was implemented after a report from the WRR22 in 1979. The report confirmed that minorities were not a temporary phase in Dutch history and that there was a need for “adequate place [for immigrants] in society on basis of political participation, socioeconomic equality and to a certain extent cultural and religious equity, to be brought about by explicit and targeted integration policies" (Penninx 2005, p. 38). This was adopted by the government as the basis for the policy. In the period, research on the position of minorities increased. A second WRR-report came out in 1989, noting that there was too little progress on the labor market for minorities, and that educational goals had not been achieved. Minority policy was said to have focused too much on cultural retainment, on rights, while obligations needed to be stressed — obligations in participation, not just on the labor market but also in Dutch society. While this report did not generate a new policy immediately, the conception of ‘obligations’ became the focus of society-in-general: today the discussions around minority ‘integration’ is very specific on their need to adapt to Dutch ‘cultural and societal’ norms, especially for migrants from a Muslim background. The last WRR-report (2001) has not had influence and has not received much attention. The govemment’s policy today can be seen as a "pick and choose 22 Wetenschappelijke Raad voor het Regeringsbeleid: Scientific Counsel for Government Policy 43 strategy" Penninx Thus the polie pollC} dt Surir. emplom There is belieie r] 20093. p. ~llathe' [ Educatim middle le 20008); T} strategy", not based on (any) research23 (Entzinger 2001; Gras and Bovenkerk 1999; Penninx 2005). Thus, while specific minority policies have been in place since 1983, the goals of the policies have not been reached, the policies have changed in focus, and today, the policy does not have a scientific or empirical base. Surinamese are (still) overrepresented in lower strata of the labor market, but their employment in middle strata has become at par with the general population (1997). There is little self-employment, “if we can use the numbers, there is little reason to believe that self-employment is an important route of economic mobility” (Niekerk 2000a, p. 24). The situation seems to be similar in education: the differences with the ‘native’ Dutch population lie in the extremes: there is overrepresentation in the lower educational tracts, and underrepresentation in the higher educational levels. At the middle level, differences have disappeared (Martens and Verweij 1997; Niekerk 2000a) Though there is a rapid rise in unemployment among Surinamese [and Antilleans] who recently immigrated the Netherlands, there is definitely a middle class emerging from within these groups, with those who immigrated before 1970 being in the most favorable social and economic position. In comparison with indigenous Dutch, however, the Surinamese and Antilleans are underrepresented in the middle and high level jobs and are over-represented among the long- terrn unemployed (Roelandt and Veenman 1994, p. 58). The position on the labor market has changed since the 19805, but it is not clear how much of a change there was. There has been a shift towards more employment in 23 With the newest Cabinet implementing all sorts of tests on Dutch norms and values, history and society, that the Dutch failed en masse last December. http://wwwteleac.nUnationaleinburgeringtest/ For some visual illustration: http://wwthanneM.com/news/special-reporg§pecial-reports- stogpage.isp?id=1964 It goes beyond the scope of this dissertation to discuss the contents of the video. 44 [hf service 5 'natite POP emplOchm bcgn in the T oierrepresei explained hf jsbclou the L‘nempl Dutch popu and the higl regardless c comparable there is a la Income is the main fill 43 perm 1998'. Gras Since it Solel}~ in th: edueated m reSuits are 1 the 370L113. 1 “I“ nor an the service sector, and labor market participation is at more comparable levels as the ‘native’ population. The more recent the migration, the larger the proportion of employment in the lower levels of the occupational structure: Surinamese that have been in the Netherlands for a relatively short period start in the lower strata. The overrepresentation of Surinamese in lower functions of the labor market cannot be explained by education levels. For all educational levels, the position of Surinamese is below the average job level for that educational level. Unemployment percentages are still three times as high as the percentage of the Dutch population. The younger groups have less of a difference in unemployment, and the higher the education levels, the less unemployment among the group. Overall, regardless of education level, unemployment among Surinamese is higher than that of comparable educated ‘native’ Dutch. Unemployment duration is substantially longer, there is a larger percentage of Surinamese unemployed for more than five years. Income levels for Surinamese are in the ‘middle category’ and income from labor is the main source for 54 percent of Surinamese men (autochtonous men: 70 percent), for 43 percent of Surinamese women (53 percent for autochtonous women) (Dagevos 1998; Gras and Bovenkerk 1999; Martens and Verweij 1997; Niekerk 2000a). Since the 19805, there is a generation of Surinamese that has had their education solely in the Netherlands, as opposed to earlier generations that were (partly) educated in Surinam. For those who have had education in the Netherlands, the results are better than for those that started their education in Surinam. The younger the group, the better the educational results, and it is estimated that educational levels will not differ significantly in the future. Of students aged 15 to 20, the differences 45 With the bachelo the age lexels i‘. with the Dutch peers are large. For example, the percentage of Surinamese in bachelor/masters programs is 3 percent, compared to a Dutch percentage of 20. For the age group of 12 to 15 year olds, there is only a slight difference in education levels (Martens and Verweij 1997): [. . .] the incorporation opportunities through the labor market have not improved. In the research of Veenman (1994) the numbers indicate that in the last 20 years, the position of allochtones compared to autochtones has not improved, and has deteriorated. This is not so much caused by the educational position of the younger generations, that position has improved compared to their parents’ position, but by the fact that educational attainments lag behind those of comparable younger autochtone generations. The intergenerational mobility of autochtones is bigger, which means that in the 19905, the discrepancies in position of allochtones and autochtones increase (Choenni 1997, p. 11, my translation). Overall, the Surinamese population in the Netherlands is a heterogeneous population in terms of occupation, education, length of stay, and ethnic origin. Surinamese are the largest minority group in the Netherlands, in 1996, the Surinamese were 2 percent of the Dutch population. In Amsterdam 10 percent of the population was of Surinamese descent (Martens 1997; Praag 2003). The population of Surinamese in Amsterdam24 was composed of the following ethnicities: Creole 70 percent, Hindustani 26 percent, Javanese 2 percent and ‘other’ 2 percent. Creole migrants are the descendants of former slaves, but this does not indicate a homogenous racial or religious group. In Surinam the main religion of the Creole population is Roman Catholic, while elements of the ‘Winti’-re1igion play a more or less important role. Hindustani are not all Hindu: the group is labeled for their origin, India, while about one-fifth is Muslim (Gowricham 2004). An interesting 2’ In a city as The Hague, for example, the percentages are different: the majority (80 percent) of Surinamese in The Hague are of Hindustani descent. 46 point. thr the lslan‘ semen v be is say Muslim Java in l: as there 1 Characte: Simi drierse E 1093}. ar the resu] point, though not completely relevant”, was made by van Amersfoort (1968): “For the Islamic Hindus [. . . ], because the Islam is more rational and can adapt easier to a society with other cultural and technical circumstances” (p.37, my translation), where he is saying that ‘Hindi Hindus’ have more problems adapting to the Netherlands than ‘Muslim Hindus’. Javanese Surinamese trace their decent to another place of origin, Java in Indonesia. Surinamese in Amsterdam are diverse on ‘race’ and religion, just as there is diversity in education, social—economic position or class, and other characteristics. Similarly, as described in the next chapters, Puerto Ricans in New York City are a diverse group. The diversity of both populations reflects the complexity of migration today, and makes conclusions based on ‘group characteristics’ doubtful. As such, if the results point towards differences in economic mobility patterns, the diversity of the groups support the importance of context. 25 Interesting in the light of recent discussions in the Netherlands, where the Islam is ‘blamed’ for a good deal of issues in society. Right-wing politicians, and the media, see the Islam as some static religion that cannot be combined with democratic values (for example Wilders, Geert. 2005. "Den Haag is laf tegen islarnistisch extremisme. Beantwoord haat en geweld van terroristen met uitsluiting en intolerantie. (The Hague responds cowardly towards islamic extremism. The answer to hate and violence is exclusion and intolerance)" Pp. 12 in NRC/Handelsblad Web- en Weekeditie voor het buiienland. Rotterdam. 47 3. PL‘ As \xit‘. York C its historical ( hate starte Netherlanc consequen The main r migrate to York socie- A bac In 149; Ben'quenv reSOUTCes‘ implemerit native DOp gifliocldea ‘ Smiling i Stan'ation I In] 3. Puerto Ricans on the Mainland As with the Surinamese position in Amsterdam, the Puerto Rican position in New York City is shaped and influenced by history, geography, and economics. The historical (colonialist) connection between the mainland and the island might not have started as early as the (colonial) connection between Surinam and the Netherlands, yet the consequences are just as far reaching. This chapter looks at these consequences linked to the economic position of Puerto Ricans in New York City. The main questions addressed are “how did the migration start, why did people migrate to New York City, and how did Puerto Ricans acquire their position in New York society”? A background to migration In 1493 the first Spanish arrived and in 1508 the arrival of Juan Ponce de Leon on Boriql’len26 marked the beginning of Spanish colonization. The island offered natural resources, among which gold, that were of interest to the Spanish. The Spanish implemented the encomienda-system: they offered 'protection' to the Tainos, the native population in return for food. This protection did not save the Tainos from genocide, of the 70.000 Tainos living on the island in 1508, only 4,000 were surviving in 1514: the brutal conditions of the encomienda-system, diseases, and starvation diminished the population rapidly. In 1509 West Africans were brought to the island as slaves. Slavery in Puerto Rico was not as widespread as in the other Caribbean islands due to the economic situation of the colonizers. The whites 2‘ Also spelled Borikén/Borinquen, the name of the island before Spanish colonization. 48 The l interman groups C1 tiff: u’ia'n. a measuri employm At the Africans l CKlSted; n 31th (A Worm), (r Petunia/u f0Finer b0 almOSt C01 denigratio Latin Ame outnumbered the people of color and the enslaved formed a small part of the total population - ranging from 10 percent in 1827 to 14 percent in 1860 to 5 percent in 1872. Visitors to the island noticed that there seemed to be greater intermingling between whites and people of color in Puerto Rico than in other places (Pe’rez y Gonzalez 2000, p. 110). The West Africans and Tainos were forced to work and live together, and intermarriage was not uncommon. Abuse by Spanish and intennarriages of the three groups created a number of 'racial' categories that were institutionalized by the expediente de sangre (1530), the Blood Registry, until 1870. The Blood Registry was a measure to exclude all that had Taino or African blood from civil or military employment. At the at height of sugar industry (1825—1845) there were about 60,000 enslaved Africans on southern coast of the island and a large variety of 'racial' categories existed: mestizo (white Spanish and Taino), mulatto (white Spanish and African), zambo (African and Taino), pardo (people with mixed ancestry and Spanish culture), moreno (mixed ancestry and African culture). Another differentiation was between Peninsulares and Criollo, the latter born in new world of Spanish ancestry and the former born in Spain (the Spanish peninsula). The class system on the island was almost completely overlapping the racial system and "It created an atmosphere of deniglation and punishment for what was culturally and physically Taino and African" (Perez y Gonzalez 2000, p. 111), "Neither Puerto Rico nor other parts of Latin America had been racial paradises" (Rodriguez 1996, p. 107) In 1521 the island's name became Puerto Rico, with the main bay of San Juan. The Spanish domination controlled the island for three centuries. Slavery was 49 abolish Li 5 dl‘ gox ernl resultec indeper l3“. Th apply. I and stll Stricken of the fi ahh051] 30m abolished in 1873, and in the last decade of the 19th century a series of events lead to US. domination of Puerto Rico. The Spanish granted autonomy to the island in 1897, and an autonomous government met in 1898, but the same year saw the Spanish-American War. This resulted in a US. 'liberation' for the island from Spain. While Puerto Ricans expected independence, the invasion by United States meant a military occupation and martial law. The island was a non—incorporated territory and the US. Constitution did not apply. Ironically, the 'liberation' by the United States meant a reduction of autonomy and still today, some of the rights that the autonomous government had, have not been reestablished. The US. domination of the island had several consequences. The economy changed radically, from dependency on crops as coffee, tobacco, sugar and cattle, to a main crop of sugar. The absentee (U.S.) owners used Puerto Rican wage laborers, and although sugar was an export crop, the total effect on import and export was negative. The currency devaluated, and when the Depression hit, it hit hard. A hurricane had stricken the island in 1928 and the combined effects created that in 1930 80 percent of the food had to be imported. Even though profits from the island were $42 million, almost 15,000 children died a year in the 19305. Some changes, as the Jones Act of 191727 which gave Puerto Ricans US. citizenship (but in 1922 it was established that citizenship did not make Puerto Rican 27 'Coincidentally’, this was also the year the US. entered the First World War. Over 20,000 Puerto Ricans were drafted. 50 part oft after th: ln tl the “CS industri depend; were fo hi gh (B Gonzali compel] Gonzale part of the United States), were supposed to improve the economic situation, but only after the Second World War and with the Cold War, real changes were made”. In the early 19505 the United States wanted to make Puerto Rico a 'showcase of the western world‘. Operation Bootstrap (Manes a la Obra) was created to increase industrialization on the island with limited success. The economy remained dependent on exports, agriculture was neglected, the 70.000 jobs that were produced were for 60 percent filled by women, unemployment of the male population remained high (Binder and Reimers 1995; Glazer and Moynihan 1963; Morales 1986; Pérez y Gonzalez 2000; Rodriguez 1989): "By 1950, one-third of the island's workforce was compelled to seek a means of survival by (im)migrating to the United States" (Perez y Gonzalez 2000, p.32). Miqration to the mainland Migration from Puerto Rico to the mainland of the United States had started before the 19505. The early stage of migration, from around 1900 to 1945, the pioneer migration, started from the villages to the northern cities in the United States. In 1910 the number of Puerto Ricans living on the mainland was estimated around 2000, "As early as 1925, Puerto Ricans in the New York colonias boasted a network of organizations and institutions designed to cope and facilitate interactions with the larger non-Latino society." (Rodriguez, Sénchez Korrol and Alers 1980, p. 1). This early migration consisted of wealthier, more urban, skilled and semi-skilled, and higher educated Puerto Ricans who predominantly moved to New York City. They 2" The 1941 Land Act was one of the changes: a maximum of SOC-acre corporate ownership of land was regulated which created some redistribution. The efl’ects were minimal. 51 iiorked they we be sligh iBinder Rodrigt settlem- hut alsc home, " mm H as \ihra labor. ii “'35 thrr Nets Tc “1“ far. turn Crez 311d Ale; Ahel aterage l migrau'm desrmaIlC 58mm h \ 3 3, liar ’hallar . SI”hate-'5 A 5” York (- worked in a vibrant industrial sector and in other regions (outside New York City) they were seasonal contract laborers. Slightly more males migrated and they tended to be slightly ‘lighter’. By 1926 there were around 20,000 Puerto Ricans in the city (Binder and Reimers 1995; Maldonaldo 1979; Ortiz 1986; Perez y Gonzalez 2000; Rodriguez 1989; Rodriguez, Sénchez Korrol and Alers 1980). In the early settlements, women played an important role, not just in adding to the family income, but also to provide information to others through contacts at work, at the bodegas, at home. The 1925 New York State Census revealed that of the 7,322 Hispanics in the main Hispanic areas”, 48 percent were women. Early research describes these women as vibrant, in both family and working life. Piecework was one of the major forms of labor, in a variety of needle and other crafts. Another way to add to the family income was through child-care and through taking lodgers: in 1925, the bulk of families in New York consisted of a nuclear family with lodgers. This household composition was far more common than a multi-family (or extended family) situation, which in turn created a 'market' for child-care (Rodriguez 1989; Rodriguez, Sa'nchez Korrol and Alers 1980; Sétnchez Korrol 1980). After the Second World War the 'great mi gration' period started. For 18 years, an average rate of 34,165 pe0ple per year left the island for the mainland. The peak of migration was in the early 19505, and New York City remained the main destination”. The people migrating in this period were younger, from agricultural sectors, had fewer skills than migrants before them, and were less educated. They 29 Manhattan's 16th, 17th, 18th, and 19th Assembly Districts in Harlem 3° Estimates vary from 34,165 to more than 42,000 a year. Tables on migration to the mainland and to New York City are at the end of the chapter 52 nere still Also. the} combined ihialdona Alers 198 organizat Durin Slmpl} 1h (211126113, 1 Culture 1‘ assumed Of'ethnic it “a were still higher educated and had more skills than those remaining on the island. Also, they were of a darker skin color than the earlier migrants. Cheap airfares combined with the prospect of manufacturing jobs enticed many to migrate (Maldonaldo 1979; Ortiz 1986; Rodriguez 1989; Rodriguez, Sénchez Korrol and Alers 1980;. Sénchez Korrol 2005). They were able to built on the institutions and organizations that had already been established in the city. During the peak migration years some scholars viewed the Puerto Ricans as simply the latest migrant group to come to New York. While they were American citizens, the argument ran, they spoke a foreign language and hailed from a different culture, thus resembling previous waves of New York, immigrants, in the end it was assumed they would pursue the same ladder of upward mobility in a traditional path of 'ethnic succession' (Binder and Reimers 1995, p. 213-4). It was a period of relative prosperity for the migrants. Jobs in manufacturing and other industries were still available. Also, this is a period of restricted migration to the US: the Quota Act of 1924 limited migration, the world wars had upset the (oceanic) streams of migrants to the United States. Job competition with other immigrant groups was limited and labor market opportunities were positive, a situation that would change dramatically in the third period of Puerto Rican migration to the mainland. The third period started around 1965, the year that also saw an important change in immigration legislation. The 1965 Immigration Act lifted the restrictions of the Quota Act and since then, a large variety of immigrants have entered the United States and New York City. For Puerto Ricans, it was not just this new job competition 53 that creal econ-om} that IS srl tor: Reu‘ migratlo: Table 3- Puerl. migrants laborers f on the is]; educallbn “here fan that created problems on the labor market, structural changes in the New York economy affected the group's opportunities. The main characteristic of this period, that is still continuing today, is a 'va y ven ’ phenomenon, a cyclical migration stream (or: Revolving Door Migration) where links to the island are maintained and migration is not as permanent as in the two periods before. Table 3-1: Puerto Rican migration to the mainland, 1900—1970 Period Net migration 1900-1920 13,000 1920-1930 42,000 1930-1940 1 8,000 1940-1950 151,000 1950-1960 470,000 1960-1970 214,000 1970-1980 65,817 1980-1990 1 16,571 1990-1999 43,926 From: Rivera-Batiz, 2005. There are over 5 million Puerto Ricans in the world, 3.5 million are in Puerto Rico, and 40 percent live on the mainland: which makes that 1.5 million are residing in North America (Guzman 1980; Pérez y Gonzalez 2000; Rodriguez 1989; Rodriguez, Sénchez Korrol and Alers 1980). Mgration to New York Citv - Puerto Rican migration to New York City follows different patterns. The first migrants were skilled laborers, but by the end of the Second World War, unskilled laborers formed the majority of the population of migrants. The economic problems on the island were aggravated by problems in the educational system: until 1949 education was in English, a second language for most. The changes in agriculture, where families lived in dire poverty without means of subsistence, and the lack of 54 industr} netisorlr eas. A 30(15. S: ordead- would 5 The few in 1 increase Dumber York [C that the Rico \V; to the u The foi next 1.mlIIl-gr (lisappe di5501i, ”ElLlhbi “eeani (leparm 10Cal 5,. industry, accumulated into changes beyond the economic: family support systems and networks were hard to maintain and people were uprooted from both urban and rural areas. As a consequence, migration increased (Grosfoguel 2003; Sénchez Korrol 2005; Smith 2001): “What they encountered was low-paying, secondary labor market or dead-end jobs, and overcrowded, dilapidated, vermin-infested tenements that would serve as their dwellings” (Perez y Gonzalez 2000, pp. 55-6). The early migrants concentrated in East Harlem, some settled in Brooklyn, and a few in the Bronx. With the (cheap) air-service from San Juan to New York migration increased drastically: from 40,000 Puerto Ricans in New York City' in 1946 the number in creased to 187,000 in 1950. By 1961 613,000 Puerto Ricans were in New York (Cruz 2005; Glazer and Moynihan 1963; Grosfoguel 2003): "Important here is that the condition of possibility for this geopolitical symbolic of showcasing Puerto Rico was to encourage the migration of the poorest sections of Puerto Rican society to the urban areas of the US." (Grosfoguel 2003, p. 109). The networks and institutions that were created by earlier migrants were an asset for newcomers, but changes in the city quickly interrupted the absorption of immigrants. With the restructuring of manufacturing, starting in the early 19605, jobs disappeared. With urban renewal, in the same period, the connections and institutions dissolved. Slum clearing and urban renewal wiped out small businesses in the neighborhoods of older (public) housing and minority entrepreneurship was negatively impacted by the rise of larger stores: "The spread of giant chain stores and department stores during the 19505 and 19605 reduced many of the advantages which local bodegas and retail stores offered within these urban communities" (Torres and 55 Rodnguez 20051. Att the labor it far orahle ( 1986. Stur Table 3-2: Rodriguez 1991, p. 260; Falcon 2005 ; Glazer and Moynihan 1963; Sénchez Korrol 2005). At the same time, less educated, more rural, and less skilled migrants entered the labor market. Overall, human capital characteristics of the group were less favorable over time, while structural conditions deteriorated (Morales 1986; Ortiz 1986; Sturz 1985). Table 3-2: Puerto Ricans in New York City, 1920-1970 Year Puerto Ricans in New York City 1920 7,370 1930 44,908 1940 61,431 1950 246,306 (187,000*) 1960 612,574 1970 811,843 From: Rivera-Batiz, 2005; * from: Glazer and Moynihan, 1963. Puerto Ricans in New York Nothing - in education, in work experiences, work training, or work discipline, in family attitudes, in physical health- gave the Puerto Rican migrant an advantage in New York City (Glazer and Moynihan 1963,p.91) In 1950, 61 percent of the jobs in the city were blue collar. A common pattern for Puerto Rican migrants in the 19605 was to start out living with relatives, and the wife and children would come to the mainland after a few years of sending money home. The working conditions were often harsh and unsafe. Between 1960 and 1980 New York lost about half a million jobs, the restructuring of the manufacturing base of New York’s economy continued in the early 19705 and in the early 19805, the decline in manufacturing was responsible for half of the total loss in jobs. In that same year, Puerto Ricans were still largely employed in 56 . manutac‘ changes higher er A S? manufacturing (33 percent of the total Puerto Rican population of the city). The changes in the working conditions in the city (and in other urban areas) required a higher educated population and a higher skill-level (Morales 1986; Smith 2001): Automation and the movement of surviving blue-collar jobs to the suburbs, the South, and to other countries have caused a sectoral decline in the number of manufacturing jobs available in New York City. Since these trends occurred more rapidly than out-migration or the retraining of blue-collar workers to fill white-collar jobs, a severe problem of blue-collar structural unemployment arose. Because of racial and ethnic prejudice, restrictive union policies, inadequate educational opportunities, and the virtual exclusion of Puerto Ricans from government employment, Puerto Ricans bore the brunt of this blue-collar structural unemployment (Rodriguez 1980a, p. 42). With the disappearance of jobs, middle classes left, and these combined strained the finances of the city. For example, in Brooklyn, the Navy yard, the Anny Terminal, the breweries, the Dodgers, all left the city in the same period. “Manufacturing plants, in particular those in the garment industry, began an exodus that increased in speed after 1970” (Binder and Reimers 1995, p. 207). The sectoral decline is a recurrent problem for the Puerto Rican working population. They were able to get a foothold in certain sectors, but when these sectors declined, it was difficult to connect to the new, growing industries. This had happened on the island between 1929 and 1939, and between 1947 and 1959. The decline in manufacturing, increased automation, and suburbanization of jobs created a similar problem as in the past. Puerto Ricans males were disproportionately represented in blue-collar work, their representation in white collar employment is smallest compared to whites and African Americans: for whites, 51 percent work in white-collar jobs, 32.5 percent of 57 African Rican m Simi exer}. le‘ mmener Puerto F account Reimers there us .last hlIt RICanS a 93mm]; H-rnHH MOS but Dot-c- mpr\ e long “.31. Afte] African American males work in white collar occupations, and only 27.2% of Puerto Rican males are employed in these sectors (Rodriguez 1980a). Similarly, Puerto Ricans are underrepresented in government employment. "At every level of the municipal government Puerto Ricans and Latinos/as are underrepresented. Latinos/as constitute 24 percent of the overall city population, and Puerto Ricans alone constitute 12 percent of the city population, yet all Latinos/as account for only 13 percent of the city workforce." (Pe’rez y Gonzalez 2000, p. 104; Reimers and Chemick 1991; Rodriguez 1980a). With the fiscal crisis of the 19705, there were large cuts in the municipal payroll, and Puerto Ricans were victims of the ‘last hired, first fired’-rule. “When the city began to hire again in the 19805, Puerto Ricans and other Hispanics benefited, but they still lagged behind other groups, particularly in management positions.”(Binder and Reimers 1995, p. 250). Housing In less than ten years [in the 19405] much of East New York was transformed from a comfortable, predominantly white, lower-middle- class neighborhood into an impoverished, overwhelmingly black and Puerto Rican area (Binder and Reimers 1995, p. 220). Most Puerto Ricans in New York lived in the barrio: segregated neighborhoods. There were some advantages, as proximity of specific stores and fellow-country men, but poverty and segregation were rampant. The New Deal brought some improvement for these neighborhoods, but federal housing was limited and there were long waiting lists. The municipal government was transformed into a ‘slum landlord’. After the Second World War, government housing improved, state and city governments became involved not only in housing for lower classes but also for 58 middle classes for being amori l implore. and d | deteriorated an | Urban renc | dorm tenement Wilding 11t’\s budding of the was labeled ',\ urban renewal. Populations th, allfimatiie hm; unafiordable ill 3005 ), After the 1. federal, State, a the borough o BIOnx and BTW Musing Units ii fUrthEr (Binder middle classes. “Of the nation’s public housing projects, New York’s had a reputation for being among the best” (Binder and Reimers 1995), p. 220). But it did not all improve, and due to segregation and poverty some of the new housing quickly deteriorated and turned into areas of a high level of crime. Urban renewal ironically did not provide new housing for the poor. “Builders tore down tenements and destroyed many small businesses as well without always providing new accommodations for the former occupants”, as happened with the building of the Lincoln Center.” (Binder and Reimers 1995, p. 220). Urban renewal was labeled ‘Negro removal’ in some cases. It is often argued that gentrification and urban renewals do not improve the living conditions of the poor (minority) populations that live in the targeted areas: these populations regularly have to seek alternative housing since their apartment buildings are converted into condominiums unaffordable for these groups (Binder and Reimers 1995; Cruz 2005; Sénchez Korrol 2005) After the 19603, the demand for housing programs has been higher than the federal, state, and local housing programs could provide for. Between 1970 and 1980 the boroughs of New York City with the highest concentration of Puerto Ricans, the Bronx and Brooklyn, were also the boroughs that lost the greatest percentage of housing units in the country. In the Reagan years, inadequate funds were cut down further (Binder and Reimers 1995; Cruz 2005; Rivera-Batiz 2005 ). 59 The I r—d .—1 ln l0 Wpulatit Poxerty : eadedl [ills resc Remark; not be e generali Ricans The Position of Puerto Ricans in New York in the 19803 Indeed, their 1980 poverty rate of 39% is actually one percentage point higher than it was in 1950. Particularly disturbing is the deteriorating position during the 19705, as the proportion of poverty rose by six percentage points (Smith 1988, p. 146). In 1980 there were 2 million Puerto Ricans on the mainland. It was a rather young population with high rates of unemployment and high rates of high school drop-outs. Poverty among Puerto Ricans was increasing and a growing number of families were headed by women. Even though the 19805 are not part of the statistical analysis of this research, it is included as an illustration of the position of both groups. Remarkably, there is little difference between these positions, a situation that would not be expected. In general, groups with a longer migration history do better, but this generalization does not seem to apply for the descriptives of the position of Puerto Ricans. Position on the labor market In the 19805, Puerto Ricans were still overrepresented in manufacturing. While 18 percent of the population of the city was employed in manufacturing, the percentage for Puerto Ricans was much higher at 28.4 percent: 60 Table ; Table 3—3: Labor market participation (in percentages)”l Group Population New Puerto Ricans York City Operators, fabricators, 15.00 27.92 laborers Precision production, craft, 8.33 9.38 repair Farming, fishing, forestry .51 .62 Service occgiations 15.08 19.77 Technical, sales, 36.34 32.54 administrative support Managerial and professional 24.64 10.19 From: Cruz, 2005. Unemployment and labor force participation "In each year the unemployment rate for Puerto Rican males in New York was about double that of whites and higher than that of blacks." (Rodriguez 1980a), p.41). The male population of the city had an unemployment percentage of 7.0, for Puerto Rican men the percentage was 11.5. For women, the city’s unemployment percentage of 7.2 compared just as negatively to the unemployment percentage of Puerto Rican women of 12.2: Table 3—4: Unemployment rates for New York City and for Puerto Ricans, 1980, in percentages New York City Puerto Ricans Men 7.0 11.5 Women 7.2 12.2 Labor force participation was lower than participation of the general population: of Puerto Rican males, 65.8 percent took part in the labor force, compared to 70.5 percent for the general population, and for women the 34 percent of female 3' The categories are not the same as the categories used in chapter 8. I have collapsed the data in similar categories for the statistical analysis. For the descriptive part of this research, I display as much information, without collapsing the categories. 61 nuncipat lienda P Educ Educ lqflnml 1970 F0 percentt mmunmi ofincrez today‘g( inincre; DID} “Umben 16 PCYCc‘ Fear 01d Table 3 participation stood against a 47.8 percent for the general population (Cruz 2005; Tienda 1993). Education Educational attainment had improved since 1960 (and 1970). For example, in 1960 only 11 percent had high school completed, this had increased to 24 percent in 1970. For women, high school graduations went from 26 percent in 1960 to 30 percent ten years later, to 39 percent in 1980. Unfortunately, "[Under these circumstances], gradual improvements in average education are no longer a guarantee of increasing income" (Hinojosa-Ojeda, Camoy and Daley 1991, p. 35). Education in today’s economy does not necessarily translate into a better labor market position or in increases in income. Drop—out rates for Puerto Rican students did not decrease since the 19705, and the numbers were higher than for the general population. "Although they constituted just 16 percent of all 16 to 21 year olds, Puerto Ricans represented 29 percent of 16 to 21 year old dropouts in 1980” (Sturz 1985, p.21): Table 3-5: Educational attainment in New York City, 1980 (percentage of persons ages 25 years and older) Educational attainment New York City Puerto Ricans Less than high school 35.77 61.92 High school 30.94 24.49 Some college 13.54 9.24 College or more 19.75 4.35 From: Cruz, 2005. 62 Income The economic slow-down or down-turn after the oil-crisis of 1973 had its consequences for every group in the labor force. But the consequences were greater for some: "all minority groups, and especially Latinos, have experienced an even greater reversal in real and relative income levels"(Hinojosa—Ojeda, Camoy and Daley 1991, p. 26). The so-called "Great U-Tum" in weekly wages had its turning point in 1973: in that year the 60 percent rise in weekly earnings that occurred since World War II changed into a 15 percent decline in real weekly wages until 1987. For minorities, the gap with white male incomes widened and the fall in real income was sharper for minority males. For females, the gap between minority and white females income became larger (while the overall gender gap decreased) and finally, the inequalities within minority groups increased sharper than the inequalities among whites. For Puerto Ricans, family income declined during the 1970s and in the early 19803 with 18.0 percent in real income (Hirschman 1988; Sandefur and Tienda 1988; Tienda 1993). The mean household incomes of Puerto Ricans were far below the mean incomes of the city. When controlled for the number of people in the household, Puerto Ricans had a mean household income per capita that was less than half of the city’s per capita mean income (Cruz 2005): 63 Table 3-6: Mean household incomes in 1980 New York Puerto Ricans Persons in Mean Mean Persons in Mean household household household household household income per income income per capita capita 2.54 $7,352 $10,802 3.10 $3,490 From: Cruz, 2005. In 1980, poverty rates were higher for Puerto Ricans than in the general population: 42.04 percent of Puerto Ricans in New York lived below poverty level, while the total of the city was 18.34 percent (Cruz 2005): “Indeed, their 1980 poverty rate of 39% is actually one percentage point higher than it was in 1950. Particularly disturbing is the deteriorating position of Puerto Ricans during the 19705, as the proportion in poverty rose by six percentage points” (Smith 1988, p. 146). Underclass? Underclass discussions have included Puerto Ricans and the "failure to participate in the labor market, along with social isolation and persisting deprivation" (Tienda 1993, p. 126) were seen as signs of underclass formation. It is difficult to see from the data since individuals are not tracked over time: thus, there might be more movement than indicated (and especially with the cyclical movement to the island). Conclusions in the formation of a Puerto Rican (part of the) underclass are not definite. “The majority of the persistent poor do not fit underclass stereotypes” (Torres and Rodriguez 1991, p. 256) In 1985, six percent of the mainland Puerto Rican population were ‘chronically detached’, which mainly means a withdrawal form the labor force. The increased instability of employment and the changes in the labor market, again, were factors 64 contributing to the detachment. Labor market behavior of Puerto Ricans is very sensitive to labor market conditions, a situation that is fairy unique because of the long migration history (Ortiz 1991; Tienda 1993). Family composition The grth of female-headed households within the Puerto Rican community was reflected in a (national) trend, but for Puerto Ricans, the growth rate was much higher. From 1960 to 1970, the percentage of female-headed families increased from 15.3 percent to 24.1 percent, an increase far larger than that of other groups. With this change, the Puerto Rican family structure has come to resemble that of African Americans. One explanation (besides rising marital instability) is that there were changes in living arrangements: the availability of welfare made it possible for these women to live on their own, independent of (ex) husband or family. From the data this interpretation does not seem to explain the increase of female- headed households. For Puerto Rican families (and black families) female-headed households gave a far worse economic position than female-headedness did for white families. It is not just the heading of the household that explains the difficult socioeconomic position of the family, other factors must play a role as well. Women in Puerto Rican female-headed households used to work more than married Puerto Rican woman in 1960 (52.8 as compared to 34.9 percent). By 1970 these numbers had changed drastically: both female heads as female non-heads had participation rates around 29 percent. Education, time of migration, fertility, or other individual characteristics cannot account for this change. The sharper decline in participation levels for female-heads could be explained by a greater dependency on 65 welfare relation desen e ln 1 all Puer present W111] ch 50 perci Figuero ESQ’TJCDH H :3" welfare, something that was not available for married women at the time. "The inter- relationship of welfare, labor-force status, and family structure among Puerto Ricans deserves further careful study" (Santana Cooney and Colon 1980, p. 82). In 1980 23.1 percent of Puerto Rican households were labeled ‘non-family’. Of all Puerto Rican families, 43.5 percent were headed by a female without a spouse present (for the total population of New York this was 26 percent). Of all families with children, the percentage of female-headed families for Puerto Ricans was about 50 percent, while for the city this was a third (Melendez, Rodriguez and Barry Figueroa 1991; Sturz 1985): The proportion of women with children under 18 and no spouse was substantial and increased between 1970 and 1980. A relatively large percentage of Puerto Rican households depended upon public assistance. Despite this aid, a large proportion of Puerto Rican families and persons had incomes below the poverty threshold in 1980 (Sturz 1985,p.6) "What may be more important is that the data show low labor force participation confined to this family type"(Sturz 1985, p. 17): Puerto Rican women in ‘married- couple households’ had a labor force participation rate of 78 percent, a percentage similar to other groups in the city. The participation rate for ‘other marital status’ is 23 percent, compared to 45 percent for the general city population (Pérez y Gonzalez 2000; Rodriguez 1991; Sturz 1985). Changes in Position since 1980 Despite the emergence of a Puerto Rican middle class, gains in education, and increasing proficiency in English, the future for many Puerto Ricans did not look very rosy in the early 19903. [...] The second generation is beginning to do better and once Puerto Ricans get up there they seem to do fairly well relative to other groups. But that’s the trick-getting there (Binder and Reimers 1995, p. 250). 66 Dur etTect o socroec mmen occupat increas: success Ofthe jt positior Socioec One int mom 0 that “PC punleti Socioec B),- 12 Pérc. IargevSi t more fe During 19803, economic growth in the United States seems to have had a positive effect on the economic position of Puerto Ricans. By the 19903 there was substantial socioeconomic progress: "poverty rates declined, welfare participation dropped, labor force participation increased - especially among women - and so did earnings and occupational advancement." (Pe’rez y Gonzalez 2000, p. 38). Educational attainment increased and people were able to move out of the manufacturing sector, but this success varied largely by geographical location, generational status, and by the level of the job skills. Overall, Puerto Ricans are in a disadvantaged socioeconomic position in comparison to non-Latino populations, and in New York City their socioeconomic status is lower than that of Puerto Ricans on the rest of the mainland. One (not tested) assumption is that those that strengthen their socioeconomic status, move out of the city (Pérez y Gonzalez 2000). On the other hand, there is evidence that “For Puerto Ricans, suburbanization and return migration were not just the purview of a growing middle class, but also of those at the bottom of the socioeconomic ladder” (F alcon 2005, p. 158). By 1990 2.7 million Puerto Ricans were living in the United States. They made up 12 percent of the Latino population of the US (estimated at 22 million), and were the largest ethnic group in New York City. The median age is 27, and there are slightly more females than males (52 and 48 percent). Table 3-7: Puerto Ricans in New York, 1980-2000 1980 860,552 1990 896,763 2000 789,172 From: Rivera-Batiz, 2005. 67 The ‘ situatior example of Puert percent occupie month a percent boarded SUbSidiz desperat (Perez .‘ Sinc their p05 Went 1 RtCan he areas Wit neigthr ireHEWal‘ hat-e nor In 19:, at 313357: The housing situation of Puerto Ricans is pretty constant over time. The housing situation in a neighborhood in the Bronx, Hunts Point/Longwood in 1991 is an example. Puerto Ricans are predominant in this barrio (and the largest concentration of Puerto Ricans is in the Bronx), the median household income is $8,448, and 70 percent of the population lives in poverty. Eighty-nine percent of the houses are occupied, and their median value is 103,000. The median rent lies around $370 a month and the average rent is 43 percent of the mean income. In the neighborhood, 7 percent of the inhabitants are occupant-owners, and 26 percent of the houses are boarded up. To put it in a different way, of the 13,011 houses, 789 are public, 828 are subsidized Section 8, and 11,395 of the houses are unsubsidized, private. Repairs are desperately needed but people often just move as soon as there is money to do so (Perez y Gonzalez 2000). Since the start of Puerto Rican migration to the mainland, until the late 19903, their position in the New York housing market has been problematic. Segregation and poverty kept them out of certain neighborhoods, and pushed them into others. Puerto Rican neighborhoods were ofien established in between white and African—American areas with little Puerto Rican contact with Anglos. The barrio started as a neighborhood with advantages as bodegas and connections, but through urban ‘renewal’ these networks were destroyed. Federal, state, and city housing policies have not improved this situation. In 1990 30 percent of Puerto Ricans lived below the poverty level (which was set at $12,575 for a family of four in that year). This was an improvement of the situation in 1980 when 36 percent of families lived below the poverty line. Incomes of Puerto 68 Ricans in New York were half of what the rest of the city made. Labor force participation of women had increased from 40 percent in 1980 to 51 percent in the 1990 Census. Their participation was mainly in low-level white-collar employment, and these gains can be attributed to the economic boom of the late 19805. Mainly because of the increase in female labor market participation, there was an increase in the median income: from $6,490 in 1980 to $8,370 in 1990. Puerto Rican men had among the lowest participation rates: 71%. Participation rates differ for Puerto Ricans born on the mainland and born in Puerto Rican: for US-born women the participation rate was 60 percent, for island born women this was 45 percent. Similarly, the participation rate for US born men was higher than that of island-born men at 73 percent (Binder and Reimers 1995; Perez y Gonzalez 2000). Even with economic growth, the unemployment rate of Puerto Ricans was still double the national average in 1990, at 12 percent. Puerto Ricans still accounted for over 20 percent of the manufacturing workforce and with the loss of those jobs, the trend towards steady unemployment did not stop. An additional employment problem came from continuous out-movement of businesses from the city. Finally, the educational system has not caught up with the changed demands in labor: Puerto Rican students were still tracked to vocational training, preparing them for the (no- more-existent) manufacturing jobs (Morales 1986; Pe’rez y Gonzalez 2000). The high unemployment rates were linked to high percentages of Puerto Ricans dependent on public assistance”. An estimate for the percentage of Puerto Ricans in 32 Including: Aid to Families with Dependent Children (replaced by national welfare program in 1996); Temporary Assistance for Needy Families; Supplementary Social Security Income; general assistance 69 New York receiving public assistance, 13.7 percent, indicates that the image of ‘all poor Puerto Ricans being on welfare’ does not hold. In fact, a larger percentage is entitled to assistance (Perez y Gonzalez 2000). In 1989, 35.6 percent of Puerto Rican families in New York City were receiving some form of assistance, while over 38 percent were living in poverty”. In 1990, 32 percent of Puerto Rican households in the US were headed by women. Since women usually make less money than men, families headed by females are more susceptible to living in poverty and in 1990 over half (52 percent) of Puerto Rican women (between ages 25-55) lived in poverty, while 30 percent of Puerto Rican men lived below the poverty line. The consequences are larger when children are involved: 42 percent of Puerto Rican children were living in poverty, fifty percent of children lived in a single-parent family. In New York City, Puerto Rican subfamilies, composed of several nuclear families or of extended family structures, increased by 182 percent between 1980 and 1990 (and for all New York City residents by 134 percent), 41,000 Puerto Rican children lived in these subfamilies (a 76 percent increase since 1980). Of Puerto Rican families, 45 percent were composed of children in a female-headed household, 58 percent of female-headed households received public assistance (Pe’rez y Gonzalez 2000; Rodriguez 1989). Educational achievements increased from 1980 to 1990. There was a sharp decline in high school drop-outs, and college attendance increased. The numbers still (Korteweg, 2005; Perez y Gonzalez, 2000. “AFDC and general assistance are commonly referred to as ‘welfare’ in the Puerto Rican community” (Pérez y Gonzalez, 2000, p. 70). 33 The 13.7% is of the total Puerto Rican population of the city, the 35.6 percent applies to households. 70 lag behind the population average, but they do reflect a positive trend in the population. These educational changes were reflected in changes in income. Still, on the other side of the story, 20 percent of Puerto Rican children under the age of 18 dropped out of high school. The differences between those that do not finish high school and those who do (and especially those who continue their education) are increasing, which might indicate a trend towards polarization among Puerto Ricans on the mainland. The situation in New York City is slightly different. Educational achievements of Puerto Ricans are very dependent of the openness of the CUNY system. Legal changes in 1976, with an open admissions policy, sharply increased Puerto Rican attendance. The fiscal crisis that followed gave a 32 percent decrease in the enrollment of Puerto Ricans between 1980 and 1988 (Binder and Reimers 1995): The recession of early 1990 gave rise in poverty level. In 1997 poverty levels were back to 36 percent and: "The 2000 census will demonstrate the effects of the economic rise and fall of the 19903 on US. Puerto Ricans" (Pérez y Gonzalez 2000, p.66). 71 4. Migration and Mobility Legally, the two groups studied are not international migrants: they moved, as citizens, from one part of the country to another. But since both groups had to cross an ocean/sea to arrive at their new destination, and since both groups had to cross cultural barriers, there is reason to compare these groups to other migrant groups. And their crossing shares characteristics with theories that explain international migration processes, that in turn help understand the position of groups in society. Thus, this chapter not only looks at theories of economic mobility in general, I also use international migration theories to understand the economic processes that describe and explain economic mobility. International Migration Historically, different groups have moved for different reasons. Only recently have ‘labels’ been attached: the difference between economic and political migrants, or refugees/asylum seekers, could only be after refugee status was defined (with the United Nations Declaration of Human Rights). Before, the boundary between economic or political migration was not defined or mentioned. That does not mean it did not exist. For example, Irish migrating in the mid-19th century were said to escape famine, and as such, they could be seen as economic migrants. On the other hand, they also escaped a colonial domination and repression by the English, which could give them the label of political refugees, if that label had existed at the time. While generalizations can be made, the labels should be used with certain caution. 72 Importantly, the motives of migration have an impact on group selection and on the characteristics of the group. Refugees are generally a different part of the population than guest workers. A prime example comes from migrants that were recognized as refugees from a communist regime: these were ofien selected from the wealthier, elite parts of society who were perceived as class enemies. Guest workers, as a contrasting example, often come from (economically) deprived regions. The comparison of these groups, with different human, social, and economic capital characteristics, is only valid while controlling for these aspects and factors (Hollifield 1992; Mullan and Majka 1997; Razin 1993; Waldinger 1989). Group comparisons on economic mobility should be done very careful, in order not to confound causes, consequences, or (causal) relationships. This chapters starts with ‘fitting’ the two groups into migration theory and migrant classifications. Migration Theorv Theories of international migration have long been based on economic assumptions. There are different bases for international migration, and contemporary theories for the start of migration patterns and for the continuation of migration incorporate these (Bean and Stevens 2003; Becker et al. 1998; Massey et al. 1998). The Start of ligation The neo-classic and new economic theories of migration are based on assumptions about economic gain. In both theories, the supply and demand of labor, and the differences in wages in origin and destination, are the main causes for international migration. If wages are higher elsewhere, people migrate, provided that the job market is fairly open (in number of jobs). 73 In the ‘new economics of migration’ more factors are included than in neo— classical migration theories. In the new economics of migration, it is not the individual by him or herself that makes the decision, but larger units, like households and families, influence cost-benefit analyses. Besides expected earnings, risk diversification plays a role in these analyses: in most developing countries, it is not only the labor market that is unstable, but also credit and capital, insurance, and future markets are highly variable and insecure. By sending a member of the household to migrate, the household diversifies its incomes and opportunities in the future. Aspects of Puerto Rican migration to the United States fit this theory: the diversification of sources of income was important for Puerto Ricans (Lewis 1965), but for the Surinamese this aspect was not central. Surinamese migration to the Netherlands started as a migration for education, and with independency, political and economic insecurity became an important push-factor, combined with a period of transitional citizenship (as explained in chapter 2). While monetary aspects are not negligible, motives of migration were not centered around the diversification of resources. Another important factor in the ‘new economics of migration’ is the reduction of relative deprivation in comparison to reference groups. The increase in income is not only expressed in money, but also in positioning the community. This latter aspect can be found in the literature for both groups. The stories of migrants that ‘have made it’ are important, and are in part a cause of the continuation, or the desire for a continuation of migration. Even when these stories are slightly more optimistic than the reality, people do believe them (Binder and Reimers 1995; Colijn and Smit 1994; 74 DeFreitas 1985; Menjivar 2000; Rodriguez, Sénchez Korrol and Alers 1980; Zhou 2002). Puerto Rican migration is difficult to fit into theoretical frameworks of international migration. The cyclical movement to and from the island seems to follow a pattern of guest-work migration. Guest workers are recruited to fill job openings that the population of a country does not want to do or when there are more jobs than people. In Europe, guest workers were recruited for industrial jobs in the 19503 and 19605, in the United States the Bracero guest work program recruiting Mexicans for agricultural work was dominant from 1942 till 1963 (Bohning 1998; Portes and Rumbaut 1996). Puerto Rican migration was largely to industrial areas in times of economic expansion, which fits the ‘European guest work’ style, but there are problems with that analogy, mainly because of the citizenship status of Puerto Ricans. The segmented labor market theory includes demand aspects of the labor market. In some sectors of the labor market, wages are not rising when there is a shortage of supply because of social and institutional mechanisms. These sectors are dependent on migration, and therefore it is from the economical organization that immigrant labor is demanded. It expands on the notion of guest work since it looks at ‘future processes’: guest work programs assumed a temporary migrant labor force, in reality, a large part of the recruited workers stayed for a longer, indefinite period of time. In the sectors that their labor was recruited for, certain jobs can become ‘immigrant jobs’ of labeled as such, which leads to the development of ethnic enclaves, a third 75 labor market sector (Bean and Stevens 2003; Dagevos 1998; Massey et a1. 1993; Massey et al. 1994; Morales 1986). Puerto Rican labor migration has aspects that match the segmented labor market theory. Their migration was not dependent on contractual labor (as is guest worker migration) since Puerto Rican movement to the US is not obstructed by legal impediments. It is a movement that is only costly in terms of money (for example air fares), not in terms of visa-issues, border-crossings, etcetera. Surinamese migration to the Netherlands follows a pattern that fits the ‘world system theory of migration’. The early migration, for educational reasons, matches colonial migration patterns, the peak of migration followed the process of independence that caused uncertainty. The economic and political uncertainties of the new country and the transitional period where no visa-requirements were implemented, caused an increase in migration. This increase cannot be explained in economic terms: it was not triggered by an increase in demand for labor, and unemployment in the Netherlands was high, not just for migrants but for all of society. Migration was at its highest during an economic downturn, in the early 19703 when the oil crisis (1973) negatively affected economies. The ‘cost-benefit-analysis’ of migration was not in terms of labor possibilities or expected earnings as the ‘neo- classical theories of migration’ would assume (Bean and Stevens 2003; Borjas 1991; Borjas 1994; Borjas 1999b; Castles and Miller 1993/1998; Massey et al. 1993; Massey et al. 1994). For both groups, the migration patterns mainly follow the assumptions described in the ‘World Systems Theory’ where an interconnected view on migration describes 76 power relations, dependencies, and disruptions on a global scale. The dependency of Puerto Rico to the US and the former colonial relation of Surinam to the Netherlands are the root for migration. The ties between core and periphery ofien originate from colonial or military roots, and these roots have had prior impacts that facilitate migration: there are cultural, linguistic, administrative, and other connections that make people knowledgeable of the core-countries (Bovenkerk 1975; Castles and Miller 1993/1998; Martin 1994; Massey et al. 1993; Massey et al. 1994; Massey et al. 1998; Staring 1999; Venema 1992; Wallerstein 1974; Wallerstein 1980). The same capitalist economic processes that create migrants in peripheral regions simultaneously attract them into developed countries. Although some people displaced by the process of market penetration move to cities, leading to the urbanization of the developing societies, inevitable many are drawn abroad because globalization created material and ideological links to the places where capital originates (Massey et al. 1998, p. 38). Continuing Migration International migration is not a ‘one-moment’ event, and the continuation of migration is important for the understanding of group economic mobility. The continuation is not random, and two theories are suggested: the social capital theory and the cumulative causation theory. According to the social capital theory, or network theory, once migration from a certain region has started, it will attract more people from that region because of the ties between migrants and the people at origin. The first migrants are able to (slowly) build up networks in the receiving countries, which enables them to create opportunities for others from their home country. 77 Because of the networks, it will become easier and less risky for the next migrants to take the step of migration. This is what leads to the institutionalization of migration: “Notwithstanding disagreement over a host of particulars, most scholars now agree that immigration is a fundamentally social process, eased by connections that link settlers to newcomers” (Waldinger 2001, p. 17). Cumulative causation theory also takes the expansion of networks into consideration, but identifies seven other causes of the continuation of migration. These causes are related to social, economic, and cultural changes in the sending countries. Through migration, changes in income, land, and labor distributions lead to displaced groups. The organization of farming does not remain the same, and well- educated people leave the country which also upsets the social structure. In the receiving countries, certain jobs will become labeled as ‘immigrant jobs’ that create more possibilities for newcomers. Finally, a culture of migration might develop in the sending country, where the behaviors might change into a ‘goal of migration’ (Bean and Stevens 2003; Massey et al. 1993; Massey et al. 1994; Massey et al. 1998; Morales 1986)”. These theories on the continuation of migration do apply to Puerto Rican migration to the US and Surinamese migration to the Netherlands. Connections in the new society were important for both groups, but the legal status of the groups made these connections less crucial than for other migrant groups. Not only did they have an easier process in crossing borders, they were legally permitted to certain welfare 3" Besides Massey et al. (1998), others use similar theories to describe contemporary international migration. For example, Castles and Miller 1993/1998; Martin 1994; Martin and Hollifield; Schoorl et a1 2000; Staring 1999. 78 benefits. The lack of legal restrictions and the availability of this ‘safety net’ might have made the decision to migrate easier than for migrants without citizenship of the receiving society. The migration histories of the two groups are shaped by historical connections. The countries of origin and of destination have ties that existed prior to the peak of migration. The links have created connections. The unequal relationship between the US and Puerto Rico, and the unequal relationship between the Netherlands and Surinam are one of the main causes of the direction of the migration. And there are some advantages of ‘colonial’ or ‘colonialist’ migration: there is knowledge about the new society, through the school system, through the political and economic dominations. For example, Surinamese children learned Dutch geography, they learned counting with tulips and apples (Bovenkerk 1975). On the other hand, these types of knowledge do not really prepare you for living in a new society, and the advantages are somewhat overrated. For the Surinamese, one important advantage is the knowledge of the language. For both groups, the recognition of diplomas gives them an exceptional benefit compared to other groups. But the disadvantages are multiple also: the long history was not a ‘neutral’ history of common goals. Exploitation and discrimination is part of that history, and although the US has ‘only’ occupied Puerto Rico after 1898, the island was colonized before and has a comparable long history of exploitation and domination as Surinam. Immigrant categories, another way to label different types of immigrants, similarly provide information about the structure of the group. Neither group came as economic or political migrants, nor did the groups came as guest workers in the 79 ‘contractual’ sense of the word. The Dutch government deliberately avoided recruiting labor from Surinam because of the realization that ‘they could not sent them back’. Puerto Ricans came to the US to do labor that others did not do (or were not available to do), but it was not contractual. The main generalizing terminology for the groups comes from theorizing on colonial/colonialist histories and migration. As such, comparisons with other groups are most appropriate when a similar colonial/colonialist connection exists. Thus, within the general theories and categories, both groups best fit the label of ‘colonial migrant’ in a World System Theory framework. This is an important generalization that makes the comparison of the two groups more legitimate. Just being an international migrant does not tell much about the underlying reasons, motivations, and causes, not does it give information about human, social, and economic capital. As I discuss throughout this dissertation, these aspects are extremely important for economic mobility. Migration grid Natives’ Opportunities One aspect of the literature on economic mobility and migration looks at the relation between wages and employment of natives and continuing migration. Wages, earnings, and employment differentials between countries are seen as a trigger for migration (Harris and Todaro 1970; Massey et al. 1998; Stahl 1995). With migration, these differences (can) decrease, wages and earnings lower in receiving societies, and employment opportunities change because of the ‘influx’ of migrants (F oner 2003; Howell and Mueller 1997; Ottaviano and Peri 2005). These notions are not supported by research: 80 A lO-percent increase in the number of immigrants decreases the average wage of natives by at most .2 percent and has little effect on the labor force participation rates and employment opportunities of practically all native groups (Borjas 1990, p. 19). Borjas does not seem satisfied with this conclusion and proposes that the wages could have been higher without immigration, suggesting that indirect effects do have an impact on the earnings of native workers (p. 24). He does not support this proposition with empirical evidence or theory. LaLonde and Topel (1992) offer an explanation for this decline in wages, and show that market conditions, not immigration, have caused a relative decline in wages for low-skilled jobs since the peak of the 19703. Immigrants 'assimilate' into the American economy, but if the economy is restructuring, this will affect their economic opportunities. Finally, Butcher and Card (1991) reach a similar conclusion: "we find no effect of an adverse wage effect of immigration" (p. 296)”. Even though research does not provide evidence that immigration has suppression effects on natives’ wages or affects unemployment levels, the assumptions remain alive in society (both in the US as in the Netherlands) and migrants quickly get the blame for unemployment in economic difficult time. Thus, the consequences might not be economic, but societal effects in (negative) imaging and stereotyping remain. Miggtion ggd nge-Pgm Besides research on the effects of migration on wages and employment of natives (Borjas 1990; Borjas 1999a; Butcher and Card 1991; LaLonde and Topel 1992; Sassen 1988; Sassen 1995a) another focus of research linking migration to the 3’ Howell and Mueller (1997) observe a decline in eamings, but leave the ‘puzzle’ open for explanations 81 mainstream labor market is on wage-disparities and changes in the disparities over time between migrants and natives (Bean and Stevens 2003; Borjas 1988; Borjas 1991; Borjas 1992; Butcher and Card 1991; Chiswick 1978; Chiswick 1979; Hirschman and Wong 1984; Schoeni, McCarthy and Vemez 1996). This literature is one-sided in that only the 'low—wage-side' of the labor market is studied. The effect of hi ghly-skilled migrants on the income distribution of highly-skilled natives is not as often researched. Chiswick (197 8) looked at foreign-born white males and their earnings. For the groups studied, income parity with native-born white men was achieved after 10-15 years. A second study by Chiswick (1979) compared three groups: economic immigrants from 'similar countries', economic immigrants from ‘other' countries, and refugees. For white males, the results of 197 8 were replicated. For the other groups, the results are more varied: "There are clear patterns of racial-ethnic group differences in economic success, even though there is substantial variation in the earnings of individuals in each group" (p. 398). The conclusion is that, when comparing foreign-bom to native-bom men of the same racial-ethnic group, the disadvantage of foreign-born disappears in a relatively short time. Hirschman and Wong (1984) look at economic attainments of Black, Hispanic, Japanese, Chinese, and Filipino men in comparison to white non-Hispanics. For the first two groups, there have been economic gains but occupational positions and earnings are still substantively below those of white men. For the Asian groups, equality in the occupational positions and earnings was reached, but this apparent equality did not reflect the higher educational credentials of Asian men: "All 82 minorities, not just blacks and Hispanics, are disadvantaged relative to the majority population" (p. 603). Hirschman and Wong use the differences in age, nativity status, residence, education, work intensity, and economic sector in their comparisons, but these variables did not explain most of the variance. Their conclusion is that direct discrimination and discrimination in distributive mechanisms, might explain the rest of the variance. Schoeni, McCarthy and Vemez (1996) modify the economic progress of Asian groups and show that the rates of progress for all groups have declined. This decline cannot be explained by the 'quality' of the immigrant groups. As LaLonde and Topel (1992), Schoeni et al. conclude that assimilation into the economy happens, but wage convergence is also dependent on the general labor market. Waldinger (2001) points out that new immigrants have lower skills than their native-bom counterparts. This conclusion increases difficulties on the analyses of general trends: most research is based on the ‘group as a whole’ and does not distinguish for place of birth or generational changes. According to Bean and Stevens (2003), these generational changes are important: research on labor market participation and mobility of the last 25 years has not shown evidence of parity in wages for the first generations. Research on second and third generations would give more insight (Bean and Stevens 2003). Borjas mainly looks at the 'quality of immigrant cohorts‘ and at the skills of immigrants to explain their wage levels: If the country of origin has more income inequality than the country of destination, the migration flow is negatively selected from the population in the country of origin. Conversely, if .the country of origin has less income inequality than the country of destination, the 83 migration flow is positively selected from the population in the country of origin (Borjas 1988, p. 96). According to Borjas, this explains the self-selection of immigrants and why today's immigrants are not doing as well as past immigrants: their 'quality‘ has declined significantly. "One factor causing systematic quality shifts across immigrant cohorts may have been the change in the quota system [. . . ]" (Borjas 1987, p. 550). To prove his point, he compares today's migrants to European migrants, looks at education and other skills and concludes that today's migrants have less skills. Borjas takes ‘national origin’ as an explanatory variable in the changes of skills of groups over generations. From his work, the conclusion is that not only do immigrants today come to the United States with less skills, the relative differences between ethnic groups do not change over time (Borjas 1987; Borjas 1991; Borjas 1992; Borjas 1994). A major problem of this approach is that he does not look at contexts: he does not include discrimination, economic restructuring, or labor demand in his discussions. Another problem is that the assumption of immigrant self-selection through international income inequalities is doubtful: Staring (1999) shows in his fieldwork in Rotterdam, the Netherlands, that the 'choice‘ for the Netherlands is not motivated by specific economic knowledge of income disparities36. The selection of migration flows is not as 'rational' as Borjas assumes. A final criticism of the notion on ‘immigrant quality’ is the lack of examination of ‘soft skills’. Human capital 3" Another way to look at the influence of international income inequalities and migration would be to compare skills of immigrants in different countries that have different levels of income inequality. The least skilled migrants should then migrate to countries with least income inequality, the highest skilled migrants to high income inequality countries. I am not aware of research that has looked at these international differences. 84 characteristics use measurements as education and work experience, but research indicates that soft skills, as motivation or the ability to interact with costumers and co—workers, determine employment. The evaluation of soft skills might not be based on the employee’s actual skills, but on the perception of the employer (Moss and Tilly 1995). Borjas' claim that 'immigrant quality' is the major explanatory factor for the lack of immigrant adaptation and assimilation in the mainstream economy today is tested indirectly in this dissertation. The groups that I study have ‘similar qualities’, if we can even apply that terminology to groups of people, but if their economic mobility is different, then other factors play a role. Economic factors and the restructuring of the receiving society, discrimination and differences in the welfare states seem to be alternative explanations. Immigrant Mobility Theories Research in the United States has a long tradition of theorizing on the aspects of immigrant economic mobility. Different models have been developed to explain the economic incorporation of (minority) groups. Predominant models of ethnic occupational mobility from the literature are first, the “ethnic succession model”, where new groups start at the bottom of the hierarchy, and others that have been in the country longer, are pushed upward. This is related to the “ethnic queue model”: groups move into the vacated positions that are opened by other groups moving upward (Lewin-Epstein and Semyonov 1986; Ortiz 1991). Both models are linked to the Dual Labor Market Theory: in both cases the new groups come to occupy the least desired positions of the labor market. There is 85 possibility to move upward, as long as the economy is not saturated. Historically, immigrants were able to improve their occupational position though the “ethnic succession model” but in today’s labor market, there are signs that positions are more permanent, with the downsizing of manufacturing and the expansion of the service industries. The expansion of the service sector gave the prediction that disadvantaged groups would increase their orientation towards this sector, particularly at the lower- wage end of the sector (Morales 1986; Ortiz 1991): "Classes do not necessarily compete with each other. The competition exists within the classes." (Morales 1986, p. 62). A third theory on minority joblessless is “skills mismatch”: skills of groups do not correspond with the job demand in the area, often the inner-city (Wilson 1996; Wilson, Aponte and Neckerman 1987). Certain groups came to the US. (and the Netherlands) for specific jobs, which disappeared with restructuring. Another aspect is that the connection between schooling and the job market ceased to match (at least for a while). The schooling and tracking of Puerto Ricans in New York shows evidence of this skills mismatch: the tracking into certain education types of Puerto Ricans students has not (completely) adapted to the new demands of the labor market (Ortiz 1991). "The viability of ethnic hiring queues, however, is related to patterns of ethnic geographical concentration" (Tienda 1993, p. 125). The weakened labor market position of Puerto Ricans in New York City is partially explained by economic restructuring and the placement at bottom of ethnic hiring queue, coupled with residential concentration in region seem to elements that explain the economic 86 position of Puerto Ricans in New York, and coupled to the failures of the educational system. A similar explanation can be linked to the Surinamese in Amsterdam. If there is a discrepancy in the economic mobility of the two groups, an additional aspect, the welfare state, might account for that difference. Surinamese came to the Netherlands in a time of economic restructuring, while other immigrant groups had already positioned themselves in certain sectors. The Surinamese came as a third main group to the Netherlands: Indonesians had migrated around the process independence of Indonesia in 1948, guest workers had entered the Dutch labor market in the late 19603. Surinamese migration increased in the 19703, and peaked around 1975. But, in 1973 the oil-crisis reshaped the Dutch economy: “Particularly, employment in manufacturing professions has decreased, as did employment in the manufacturing industry" (Veenman and Roelandt 1994, p.35). Their educational levels were lower than ‘Dutch’ educational levels, and their inadequate qualifications did not place them in a competitive position. The (ethnic) hiring queue ‘crowded out’ the lower educated, and selection practices by Dutch employers added to the queueing of Surinamese (Martens and Verweij 1997; Niekerk 2000a; Veenman and Roelandt 1994). The concentration of Surinamese in the Bijlmermeer created another problem: the Bijlmermeer was supposed to attract businesses, which did not follow the expectations. The creation of a labor market around the new built housing in the area did not occur (Klerk and Amersfoort 1988; Veenman and Roelandt 1994). 87 Mobility and Labor Market Context Immigrant economic mobility is not just explained by ‘human capital characteristics’ as (average) education, time of arrival, length of stay, work experiences, as theories on ethnic queueing and skills mismatch indicate. Human capital characteristics cannot explain the position of Puerto Ricans in New York in a satisfactory way. For example: Although a substantial literature documents the importance of education for labor market success, the low educational achievement of Mexicans challenges the completeness of the human-capital explanations. Mexicans have not experienced declines in labor market standing and economic well-being comparable to those of Puerto Ricans, even though their educational levels are similar (Tienda 1993, p.124) There is no linear relationship between education and employment, the relation between educational achievements and employment status for different groups (and sub-groups) in the United States varies (Tienda 1993; Torres and Torre 1991): [. . . ], the human capital approach translates into a myopic view of policy changes that might contribute to a change in income disparities for the Latino population. It provides a narrow view of production and an even more limited understanding of social reproduction in the economy (Torres and Torre 1991, p. 280). Contextual factors are important in the economic adaptation and incorporation of immigrants. Labor market characteristics shape the opportunity structure of a country or of a city. For example, the structure of the United States’ labor market at the peak of migration from Europe (around the turn of the 20th century) had many unskilled at the bottom, a small number of workers in the middle, and very few on the top of the income hierarchy. The supply of workers, the demand for labor, and the native population was very similar to the immigrant population at the time. Between 1924 88 and 1965 the shape changed and refugee-migration supplied workers at the top while the Bracero-program supplied labor at the bottom. And this dichotomy of top- and bottom-end work supply remained very strong after 1965, a trend strengthened by the 1990 Immigration Act. The demand for top-end workers has increased. "1. . .1 continued immigration appears to contribute both to the maintenance and expansion of a polarized economic system, while also lowering the costs of operating such a system" (Sassen-Koob 1985, 305). The situation in Europe was different: before the 19603 labor-demand was concentrated on unskilled work or for the secondary segment of the labor market. Supply for the higher end of the job distribution was provided by the native population and by migration from OECD37 countries. In the 19903 the demand for skilled labor increased but there is no dichotomy of demand and supply as in the United States (Bohning 1998; North 1978). Sassen (1988) sees the restructuring and global expansion of the capitalist economy as crucial in changes in migration since 1965, a theoretical view in line with World Systems Theory. The contemporary situation in global cities can be summarized: Immigration can be seen as providing labor for: (1) low-wage service jobs, including those that service (a) the expanding, highly specialized, export-oriented service sector and (b) the high-income lifestyles of the growing top level professional workforce employed in that sector; (2) the expanding downgraded manufacturing sector, including but not exclusively, declining industries in need of cheap labor for survival, as well as dynamic electronics sectors, some of which actually can be seen as part of the downgraded sector. A third source of jobs for immigrants, is the immigrant community itself (Sassen 1988, p. 22). ’7 Organization of Economic Cooperation and Development 89 The strength of Sassen's argument is that she looks at both the political and the economical context to explain migration. An example of the impact of political decisions is the 1965 Immigration Act in the United States. The Act made provisions for entry for highly skilled immigrants, who enter the 'primary labor market' (or: the monopoly sector, (O'Conner 1973), where jobs are stable, relatively well-paid, and good working conditions. Low-skilled migrants do not have the possibility of employment in the primary labor market, but are dependent on either the secondary labor market (or: the competitive sector, (O‘Conner 1973), where characteristic opposite of those of the primary labor market prevail) or the ethnic economy, a third economic sector (Wilson 1980)”. A fourth sector in the contemporary economy is the 'transnational sector': "transnational entrepreneurship lies at the intersection of immigrant enterprise [. . .] and the broader field of transnationalism" (Portes, Haller and Guamizo 2001, p. 7-8). Finally, the expansion of government related employment is a fifth economic sector (O’Conner 2001). Another point of strength in her work is the focus on global cities: regions are affected differently by globalization and migration (Melendez 1991; Ortiz 1991; Rath 2002b; Sassen-Koob 1985; Sassen 1988; Sassen 1995a; Waldinger 1989; Wilson 1996). The changes in New York City and in Amsterdam are the topic of chapters 2, 3, and 7. 38 The informal economy can be considered a fourth sector of the economy. There is evidence that this sector is growing in Western societies (Light, Ivan, and Steven J. Gold. 2000. Ethnic Economies. San Diego: Academic Press, Portes, Alejandro, and Saskia Sassen-Koob. 1987. "Making it Underground: Comparative Material on the Informal Sector in Western Market Economies." American Journal of Sociology 9330-61., and comparisons with the secondary sector have been made. 90 The overlap between immigration and economic mobility in a World System is important. The Dual Labor Market does not just look at immigrants on the labor market but at all types of minorities. Since Puerto Ricans and Surinamese are citizens of the receiving countries, their position is different than the position of other migrants. One of the ‘key notions’ of the Dual Labor Market is that migrants end in the second tier because they do not have citizenship rights. It is easier to hire and fire them at random (Hollifield 1992). The segmented labor market allows for more economic sectors. Segmentation is not just a dual division in the labor market but multiple sectors, or segments, can exist, such as ethnic economies and government sector. The dual or segmented labor market includes a variety of contextual aspects. Discrimination is part of these aspects, on different parts of the labor market: at the entry level, in the process of job allocation, and for differential wages. The framework of the dual labor market focuses on minority concentrations in the secondary market: Although discrimination against non-white workers does and can occur in the primary labor market, this problem is of lesser importance than the problem of concentration of non-white workers in the secondary labor market - a problem deeply rooted in the subculture of non-white workers and the structure and nature of the secondary labor market (Torres and Torre 1991, p. 274). I focus on three of the mentioned sectors: the primary and secondary sectors and government employment. The latter has been important in the economic advancement of minorities, and is connected to the welfare state (but not synonymous: not all that work in government employment work as part of the welfare state). Ethnic economies and the transnational sector are excluded from this research because the groups do not have a significant proportion of their population working in these sectors. 91 "About two thirds of the Latino labor force is located in the peripheral industries and more than 60% is concentrated in the low-wage occupational segment" (Torres and Torre 1991, p. 276). The overrepresentation in the low-wage occupational segment of Latinos is not just due to discrimination, human capital characteristics also play a role, just as industrial policies, the structure of the capitalist economy, and general aspects of (immigrant) economic mobility. Economic mobility History affects the results of mobility analyses by altering the occupational/industrial structure and by causing what amounts to exchange mobility (Miller 2001, p. 314). Economic mobility is not a topic solely linked to minority groups in the ‘racial/ethnic’ sense. Economic divisions are almost as old as society itself, and the study of these divisions is as old as sociology itself. The literature on economic mobility traditionally focused on (white) male income; in the 19603 other (male) groups were included, and not until recently research expanded to include data on female mobility. Research in the (early) 19603, a time of economic boom while migration was emerging but not a topic of interest, mobility questions in the US. focused around the occupational inheritance and the inheritance of poverty (Blau and Duncan 1967; Duncan 1968; Gans 1968; Lewis 1968; Lipset and Bendix 1967). With the War on Poverty and the Civil Rights Movement, interest in the opportunities of minorities increased in the US. In the Netherlands, minority studies were still on religious groups (Protestants, Catholics) and on the anti-socials, and only in the 19803 did 92 research on other minorities emerge (Duncan 1966; Matras 1980; Miller 1960; Nakao 1992; Rath 1999; Wusten 2001). The focus in more contemporary research in economic mobility is still comparative, as in the past, but comparisons have been taken to an international level. Not only are comparisons taken from the mobility patterns of different class levels to other groups, as immigrants, comparisons of patterns of different countries are well- studied today. Importantly, these comparisons have revealed aspects important to economic mobility that are similar to the aspects distilled from the literature on immigrant economic mobility. Definitions In the 20th century, economic stratification, class, and economic mobility were topics at the forefront of social studies. The terminology linked to economic mobility is complex: occupational mobility, social mobility, economic mobility, class mobility, increases in income, movement out of poverty, social mobility, these were all connected to what could be described as “movement of individuals among positions defined by the structure of the division of labour” (Ericson and Goldthorpe 1992b, p. 1). Mobility rate is not yet as standardized a measure as, say, crude birth rate; nor are there defined mobility rates as comparable and widely accepted as any of the refined birth rates (Matras 1980, p. 411). In the literature, there is great overlap between social (status) mobility, economic mobility, class mobility, and occupational mobility. “In more general terms, a person who raises his occupational status will normally seek also to raise his social status” (Lipset and Bendix 1967, p. 6). Occupational mobility and social mobility are not the 93 same, there are multiple aspects of social mobility, as education, income, power, knowledge, consumption, or in general, access to resources. While these aspects are important, most research equates social and occupational mobility, if only just for the reason of measurement (J onsson and Mills 1993; Lipset and Bendix 1967; Matras 1980) Economic mobility, in its narrowest definition, would be income mobility. In comparative research: “Income attainments are difficult to measure and even more difficult to compare; few comparative mobility studies incorporate income attainment measures” (Matras 1980, p. 418)”. Occupational mobility is used instead: Studies of social mobility have traditionally focused upon occupational ranks as reflecting the overall social ranks of individuals; on occupational categories or groups — usually ordered hierarchically - as reflecting significant strata, status groups, or classes; [and on intergenerational occupational mobility, career mobility, and marriage of women of given occupational origins to men of given current occupational ranks, as representing social mobility] (Matras 1980, p. 407) Class mobility differs from occupational mobility by combining different occupations into classes: We combine occupational categories whose members would appear, in light of the available evidence, to be typically comparable, on the one hand, in terms of their sources and levels of income and other conditions of employment, in their degree of economic security and in their chances of economic advancement; and, on the other hand, in their location within the systems of authority and control governing the processes of production in which they are engaged (Ericson and Goldthorpe 1992b, p. 40). 39 Similarly, for international comparisons, occupational prestige is more comparable than ‘income’ as a sole measure of social or economic mobility. The problems of measuring ‘income attainment’ are gravitated by the problem of using both income and prestige: these aspects are highly correlated (.88) (Treiman 1977). 94 International Economic Mobility Since World War Two, different ‘generations’ of economic mobility research are identified. The first generation of research looked at national occupational prestige inquiries, and the international comparative focus was between industrialized and non-industrialized countries. The notion was that industrialization would create greater openness of societies, with a pattern towards convergence. Their conclusions were rejected by later generations of research. The second and third generations of mobility research improved the standards of data collecting. New scales were developed and new methods of analysis. Path analysis was introduced by the second generation, the third generation of mobility research replaced multivariate linear models by loglinear modeling. The CASMIN—project compared intergenerational mobility patterns in 13 industrialized countries. From the project, the standard classification of occupational classes for comparative research emerged. The core model was a loglinear model with levels that revealed interpretable parts of mobility: as inheritance effects, sectoral effects, hierarchical effects, and affinity effects. It also revealed country-specific patterns and large between country variations in mobility (Ericson and Goldthorpe 1992a; Ganzeboom, Treiman and Ultee 1991; Matras 1980; Treiman and Ganzeboom 2000). The fourth generation of mobility research included the impact of institutional arrangements. Multilevel analysis was used to look at factors as social development, the political structure, and levels of educational and occupational reproduction. For example, “They showed that vocational education yields higher returns in countries where vocational training is occupation specific” (Treiman and Ganzeboom 2000, p. 128). 95 Comparative Differences: MobilityRegmes From the outset, cross-national and cross-temporal comparisons have had a central role, since such comparisons provide the only way to determine whether, to what extent, and in what ways the intergenerational transmission of advantage is dependent upon other aspects of social organization, and what its consequences are (Ganzeboom, Treiman and Ultee 1991, p. 278). The first researchers in an international comparative perspective looked at the differences in ‘openness’ of industrialized and non-industrialized countries. The assumption from Liberal Theory was that with industrialization the openness would increase. The influence of father’s occupation decreases with industrialization, education becomes more important, and parents’ status is of less importance for possible educational levels. "Free education is more readily available in industrialized societies and hence in particular is more readily available to those from low status origins" (Treiman and Yip 1989, p. 376; Ericson and Goldthorpe 1992b; Goldthorpe 1980/1987). A second assumption was that industrialized countries would show a convergence in mobility or openness. A debate on convergence through industrialization is centered around the difference between structural and circulation mobility"). Structural mobility is "that part of observed mobility which is necessitated by changes in a country's economic structure", circulation mobility the "mobility net of structural mobility, the kind of mobility that indicates genuine openness" (Ultee and Luijkx 1986, p. 193). Structural mobility is caused by changes that create a different occupational structure, as happened with the transition from an industrial society to 4° There are several labels to these two forms of mobility: structural mobility is also called ‘forced’ mobility or ‘competitive balance’, circulation mobility is equal to exchange mobility or ‘social fluidity’ (Ericson and Goldthorpe 1992b; Lipset and Bendix 1967; Matras 1980; Ultee and Luijkx 1986). 96 the post-industrial (service) economy. These changes were considered to have a different impact on opportunities than changes in circulation mobility. Circulation mobility is measured through the changes that reflect a similar labor market structure and movement of people within that structure. It shows the openness of society through mobility within a certain structure, instead of movements caused by structural change as happens in structural mobility. While the discussion around structural and circulation mobility is interesting, the differentiation and its ‘single valued characterization of openness’ has been partially abandoned (Ericson and Goldthorpe 1992b; Matras 1980; Miller 2001; Ultee and Luijkx 1986). “I But], as regards attempts to distinguish between structural and exchange mobility per se, [. . .] these are now outmoded and unhelpful and should no longer be pursued” (Ericson and Goldthorpe 1992b, p. 59) and "It is desirable to focus on structural and circulation mobility in one investigation" (Ultee and Luijkx 1986, p. 194). Empirical support for the assumptions that were generated through Liberal Theory on the consequences of industrialization on mobility is inconsistent, and other factors besides industrialization play a role“. A second aspect (or cluster of aspects) that has an impact on mobility is the political constellation of a country. Political influences, and specifically welfare state measures, have proven to affect mobility significantly. With the fourth generation of 4' The Liberal Theory was not the only theory postulating similar trends across industrialized societies. The Lipset and Zetterberg (LZ hypothesis): there is no association between mobility rates and rates of economic growth: “our tentative interpretation is that the social mobility of societies becomes relatively high once their industrialization, and hence their economic expansion, reaches a certain level” (21): the threshold of mobility is at a fairly early stage of industrialization, and similar across industrialized societies. A reformulation of the LZ hypothesis was suggested by Featherman, Jones and Hauser (FJH hypothesis), which was also not supported by empirical results (Ericson and Goldthorpe 1992b). 97 mobility research, two sets of (similar) political factors were research: 1) differences between rightist versus more left governments and 2) differences in mobility patterns between types of welfare state. Both structural as circulation mobility are affected by exogenous forces (as industrial development) but also by regimes: more lefiist regimes for example, have more circulation mobility: the structure is more ‘genuinely open’ (Ganzeboom, Luijkx and Treiman 1989; Ganzeboom, Treiman and Ultee 1991; Matras 1980; Treiman and Ganzeboom 2000; Ultee and Luijkx 1986). The welfare state, as a system of (re)stratification, affects opportunities in society. One way the welfare state shapes opportunities is through adjustments to structural change: Because the structure of institutional constraints and inducements differs across nations, we expect that nations will differ in the extent to which institutions shield or channel the impact of structural change [on individual outcomes] (DiPrete et al. 1997, p. 320). The welfare state has different ways to influence stratification of society, and this is not the only way of interfering with the labor market, the economy, and mobility of individuals and groups (DiPrete et al. 1997; DiPrete and McManus 1996; Esping- Anderson 1990). _I_-'_actors on Mobility The search for factors that have an influence on mobility (intergenerational) has given the same conclusion repeatedly: education is one of the main aspects for intergenerational mobility in industrial societies. It is more important than father’s education or father’s occupation, aspects that play an important role in pre- industrialized or industrializing countries (Blau and Duncan 1967; Ganzeboom, 98 Luijkx and Treiman 1989; Ganzeboom, Treiman and Ultee 1991; Graaf and Luijkx 1994; Lipset and Bendix 1967; Matras 1980; Treiman 1977; Treiman 2001; Treiman and Yip 1989; Ultee and Luijkx 1986). The ‘openness’ of educational opportunities in industrialized societies is something up for debate: some conclude that education is independent of social origin, others link educational achievement to the social, financial, and cultural capital of the family. The latter seems to be supported more. “Those who come from families with high cultural capital will tend to obtain more schooling than will others [. . .] do better in school, [...] begin schooling with a head start” (Treiman 2001, p. 307; Ultee and Luijkx 1986). Financial capital affects educational attainment in three ways: it pays for schooling, it can pay for a higher quality of schooling, but it also can give access to better schooling through housing in better school districts (in the US). Social capital has a relationship with educational attainment through aspirations and expectation -— or parents, of peers, and of teachers (Treiman 2001). On the one hand, educational opportunities are affected by the status of parents, on the other hand, there is a loose connection between the socio-economic status (SES) of parents and their children: fewer than 10 percent and no more than 25 percent of the variance in SE8 is shared by parents and children. Fewer than 10 percent of the children have the same occupation as their father (Treiman 2001; Treiman and Yip 1989). Thus, educational achievement and occupational status are central to social status in the modern world, and their determinants, correlates, and consequences, including 99 their relation to each other, within and across generations, are the primary foci of social stratification research (Treiman 2001, p. 298-99)”. Changes in the occupational structure, society’s development, political organization, levels of inequality, and human capital characteristics of individuals and groups, these are all aspects or factors that play a role in mobility (Graaf and Luijkx 1994; Matras 1980; Treiman 2001; Treiman and Yip 1989). In chapters 7, 8, and 9, I have integrated these factors, and others, while comparing the different contexts of reception. Occupational Scales Four scales are accepted to measure occupational categories. International Prestige Scales have proven to be robust over time and place. Secondly, the measurement of socio-economic status (SES) is accurately done through the Social Economic Index (SE1) developed first by Duncan, based on a combination of average education and income of occupations, and adapted by Ganzeboom to the ISEI, the International Social Economic Index: “The crucial difference between prestige and socioeconomic status is the position of farmers. Whereas farmers enjoy about average prestige around the world, they tend to be near the bottom of socioeconomic status scales such as Duncan’s SEI”43 (Ganzeboom, Treiman and Ultee 1991, p. 283; Dessens et al. 2003). “2 The relation between social origin and social mobility of the successive generation (or between educational achievement and parents’ SES is not clear: on the one hand, the relation is described through cultural, financial, and social capital, on the other hand: 'Because education is largely independent of social origins, the consequence is that occupational status attainment is mainly a matter of achievement and not of ascription’ (Treiman and Yip 1989, p. 392). What is clear is that education is an important factor in social mobility. ‘3 This study looks at city populations only, which makes the ‘farmers question’ irrelevant. 100 The third measurement of occupational classes for international comparisons came from Ericson, Goldthorpe, and Portocarero (1979). This scale (EGP scheme) was based on class-characteristics of occupations and includes different aspects. The first is the sector of the labor market: nonmanual workers, manual workers, or farm workers. A second type of attributes came from employment: self-employed or salaried. Third, skill levels were used to rank manual and nonmanual occupations into three strata, and fourth, supervisory status was included: how many were supervised (for supervisors and managers) or how many employees (for self-employed). This classification was adapted by Ganzeboom et al. (1989)(Dessens et al. 2003; Evans 1992; Ganzeboom, Luijkx and Treiman 1989; Matras 1980; Treiman 2001). The fourth way of scaling occupations is through the International Standard Classification of Occupation assembled by the International Labour Office (DiPrete et al. 1997). In 1957, the first version of the ISCO was developed, and adaptation were made in 1968 and 1988 through International Conferences of Labour Statisticians (9m, 11‘“, and 14th ICLS)“. The versions are the base of classification of the national labor market in many countries, and while national and regional differences exist, the classification allows for international comparisons”. The ISCO—88 is a tool for organizing jobs around the variable ‘the main tasks and duties of work performed’46. As such, it is organized around skill-level and skill- specialization, and overlaps with two of the criteria of the other classifications of tTable 4-1, industry and skill-level. In the table, the ten major occupational groups are 4" From: http://www.ilo.org/public/english/bureau/stat/isco/index.htm 4’ From: http://www.ilo.org/public/english/bureau/stat/iscofrntro3.htm ‘6 From: http://www.ilo.org/public/english/bureau/stat/isco/docs/introS.htm 101 collapsed on their skill-level, to illustrate the similarity in categorization. In the model, both skill—level and type of industry are used: both the Census and the SPVA use categories closely following the ISCO-88: Table 4-1: Occupational categorizations 10 category EGP (1979) 1 large proprietors, higher professionals and managers II lower professionals and managers 111 routine nonmanual workers IVa small proprietors with employees IVb small proprietors without employees V lower grade technicians and manual supervisors VI skilled manual workers VIIa unskilled and semi-skilled manual workers IVc Self employed farmers VIIb (unskilled) agricultural workers Ganzeboom, Lufikx, Treiman, 1 989 ISCO—88 by skill level: 1 Legislators, senior officials, managers 2 professionals 3 technicians and associate professionals 4 clerks 5 service workers and sales workers 7 craft and trade workers 8 plant and machine Operators and assemblers 9 elementary occupations 6 skilled agricultural and fishery workers Armed forces From: Ganzeboom, Luijkx and Treiman, 1989; http://www.ilo.orgjpublic/english/bureau/stat/isco/iscoSS/publ3.htm 102 Two issues come to mind in using these categories: the lack of measurement of wealth, and questions on gender“. Wealth is generally not included in mobility research, something that is problematic to a certain extend. On the one hand, income from wealth greatly affects opportunities, on the other hand, one could argue that those are not completely linked to the opportunity structure (Ganzeboom, Treiman and Ultee 1991). Gender issues arise not only from the lack of earlier studies on female mobility, but also from measurement problems: whose mobility and status are measured? With higher female participation rates in the labor force, research on opportunities for women is crucial. The class position, as measured by occupational categories, of an individual might change by marriage or some other form of communal household. Women’s or men’s occupational status might not reflect her/his class status because of the shared household. To measure the class position of an individual thus causes problems. On the other hand, if some ‘average class position’ of the household would be taken, this needs to be controlled for number of dependents, number of people in the labor force in the household, etc. It seems that the least problematic ‘common’ measure would be by taking the occupational status of the head of the household. Overall, there seem to be problems, especially on an individual measurement level, but also on combined measurements (DiPrete 2002; Matras 1980). I have avoided this problem by looking at the group as a whole: due to limits imposed by the data sets (mainly the ‘short’ immigration period), I cannot look at ‘7 A third issue would be the ‘perception’ of opportunities: the US. opportunity structure is not that different from European structures, but the perception of endless possibility creates higher expectations (Matras 1980). 103 intergenerational or intragenerational mobility. I look at the mobility of the groups to see if their position in society is different over time, and to see if there are differences in the two welfare states of destination. Conclusions The patterns of migration are different than patterns of other groups, although the general pattern of decolonization and migration is common. The initiation and continuation migration for these specific groups is mainly different in one consequence: a very diverse p0pulation migrated. Unlike guest workers, who often came from very concentrated regions of the sending country, Puerto Ricans on the mainland and Surinamese in the Netherlands came from similar (class) parts of society (Mollenkopf 2000; Rath 2000). In their diversity it is difficult to rely on ‘general characteristics’ or ‘typical migrant patterns’. The groups need be analyzed not only as immigrants, but also in relation to different theories, and especially in relation to general economic mobility trends. From the literature, three sets of characteristics are relevant to economic mobility: the composition of the group (or human capital characteristics); the conditions on the labor market (segmented labor market and World Systems; government interference (the welfare state in different forms). These three sets are defined by Portes and Rumbaut (1996): “For immigrants, the most relevant contexts of reception are defined by the policies of the receiving government the conditions of the host labor market, and the characteristics of their own ethnic communities” (p. 84). I discuss these contexts in chapter 7, but first, the next chapter discusses discrimination and segregation, aspects of minority incorporation that are not included in the model but 104 cannot be ignored. Discrimination overlaps with at least two of the contexts, and cross-sects society and the reception of new groups in many different ways. The chapter is followed by a chapter on the welfare state. 105 5. Discrimination and Segregation Racism and discrimination, and the consequences, is widely studied in the US. Knowledge about these practices is generally (within the social sciences) accepted, and a statement on racism or discrimination in society does not raise many eyebrows. In the Netherlands, research tends to focus on the 'limitations' of migrants or minorities instead of on the limitations of society (Bovenkerk 1978/1979b; Choenni and Cain 1995; Essed 1991). "The Netherlands has a poor record in the field of academic studies into race and ethnic relations" (p. 511) and "studies have tended to concentrate their search for racists in Dutch society among right-wing extremist groups and parties, in an effort to purify themselves and other whites" (Choenni and Cain 1995, p. 513). Case studies (for example Marger 2003; Wade 1997; Winant 2001) reveal that although there is a ‘world racial system’ (Winant 2001, p. 137), countries and regions have different applications of racism and discrimination. Time, historical change, is another aspect of diversity in racial and ethnic categories that is discussed in this chapter. Historical notions of biological inferiority are not as prevalent today, cultural racism could be seen as ‘ethnicism’ and as such loose its negative stigma. The consequences of ethnicism are similar to consequences of racism, where the ‘other’ is defined by group characteristics and stereotypes (Omi and Winant 1986/1989; Vasta 2006; Winant 2000; Winant 2001). 106 Race and Ethnicity in the US The racial formation in the US is strongly determined by the history of internal slavery. The ‘one-drop-of-blood-rule’ has made the bifurcation of ‘race’ in the US extreme: the US is considered color-blind by several Latino scholars, in that they do not recognize the ‘shading’ of people. In the US, one drop of African ancestry gives a racial categorization based on that drop, and not on any other heritages (Gold 2004; Lee 1993; Omi and Winant 1986/1989; Rumbaut 2000; Vasta 2006; Winant 2000; Winant 2001; Wright 1990). For Puerto Ricans this meant that "[They] entered a US. society that had a biologically based, biracial structure and that had tended to accommodate multiracial cultural groups into this structure" (Rodriguez 1996, p.105). The biracial division of society in the early period, when there were little Hispanic migrants in the northeastern part of the US, gave: She asked my nationality and my cousin answered ‘Puerto Rican’, but she wrote down ‘Negro’. My cousin protested, ‘No, no, no, not Negro, Puerto Rican.’ She gave him a look but she erased ‘Negro’ and wrote down ‘Puerto Rican. ’ That was my first experience of that kind up here (Lewis 1965, p. 227). Today, this has changed, and the category ‘Hispanic’ is a new label that facilitates categorization of people in certain groups (Rumbaut 1996; Waters 1996). But generalizations are problematic. In the literature, comparisons and similarities between Puerto Ricans and African Americans are made (Massey and Denton 1993): both groups are citizens of the United States, both groups were among the ‘early non-white populations’ in cities of the northeast. There are aspects that are similar, but Puerto Ricans were raised in 107 Spanish, and Puerto Ricans have the connection to the island. Finally, and crucially, identification is different: the racial hierarchy on Puerto Rico is not determined by the one-drop-of-blood-rule that defines someone as ‘black’ in the mainland-US. context, but it is more fluid and there are more categories. A person could encounter his or her ‘blackness’ for the first time when landing on mainland-soil. The fluidity of the Puerto Rican ‘racial categories’ is at least in part determined by the (colonial) history. Identification is questioned in a second way: It raises the question of why they would expect darker Puerto Ricans to identify with blacks. It appears that the surprise is the result of applying the North American racial classification standards; thus, because you look black, you must be black and to be black is to be just like American blacks, not to have a different culture or language. Although it is possible and easy to distinguish between Italians, Greeks, and Germans, it is not possible, or perhaps desirable, to distinguish between African blacks, West Indian blacks, and American blacks. Culture becomes subordinate to race, and perhaps in the case of blacks, it ceases to exist altogether in this conception (Rodriguez 1996,p.118) Identification is cultural or national, with the island, unlike on mainland US. "This is not to say that Puerto Ricans feel no racial identification, but rather that cultural identification supercedes it" (Rodriguez 1980b, p. 26; F alcén 2005). This also implicates issues with the categorization with other groups into ‘Hispanic’ or ‘Latino’. Puerto Ricans are the only Hispanic/Latino group that came to the US as citizens. They also were one of the first groups that migrated, which made the ‘racial’ experience different, but also gave a different labor market start: Puerto Ricans came at a time where there was a demand for unskilled manufacturing labor, jobs they took, and jobs that have been downsized and that have disappeared ever Since. 108 Although discrimination may still be a major factor accounting for the disadvantaged status of Puerto Ricans, [it does not address the issue of why the economic status of Cubans and especially Mexicans has not followed suit]. While not denying the importance of prejudice in maintaining socioeconomic inequality along racial and ethnic lines, a structural interpretation is consistent with the uneven regional effects of economic growth and decline that occurred during the late 19703 and early 19803 (Tienda 1993, p. 135). As with the grouping with African American, identity issues are not respected in these larger categories, but more importantly, the history and specific problems of a group are ignored. This does not mean that Puerto Ricans do not face the same or similar issues as African Americans or as other Latino groups. There are too many differences between Puerto Ricans and Afiican Americans to consider these groups together. On the other hand, there are strong warnings against 'lumping' all Hispanic groups together, and to carefully analyze national origin groups because of the peculiarities of each group through each migration history and context. It means that theories on other groups (and importantly, on racial segregation) do apply, but that one always needs to keep the group characteristics in mind. Also, it does not mean that the position of Puerto Ricans in mainland-US society is not influenced by the US color line. The migration of Puerto Rican was earlier than the main migration of other Spanish groups (with the exception of Mexican presence and migration in the United States (Haslip-Viera and Baver 1996), and especially darker Puerto Ricans were forced to move in ‘black’ neighborhoods. Puerto Ricans are in a unique position in US society. On the one hand, there are similarities with the African American position, on the other they are a migrant group from a different culture with a different language (Haslip-Viera and Baver 1996; 109 Perez y Gonzalez 2000). And unlike most other migrant groups, Puerto Ricans have US. citizenship upon arrival. Race and Ethnicity in the Netherlands [..] Over the centuries the Dutch preacher and salesmen mentality has enabled the Netherlands to gloss over racism where it existed, while ensuring impediments to conducting business were removed (Choenni and Cain 1995, p. 513). The Dutch have been able to hold on the notion of being 'tolerant' for a long history: migration to the Netherlands is not just a contemporary process. In the 16th and 17’h centuries, Jews, Huguenots, and other religious groups found refuge in the Netherlands. German labor migration existed in cyclical periods for harvests, and in relation to both world wars, refugees were welcomed in the Netherlands (Beer 1998; Comelis 1990; Heelsum 1997; Klerk and Amersfoort 1988). These ‘streams’ of migrants created a long history of a ‘tolerant nation’. Indeed, the groups were welcomed, but some group characteristics need to be taken in consideration: the early migrants were all fairly well-to-do, and contributed to the Dutch economy. Also, there was extensive out-migration, and the lines of ‘foreign’ and ‘Dutch’ were blurred. After World War H, migration patterns changed. A first migration wave to the Netherlands was from the former colony Indonesia, but net-migration shifted towards a positive number in a second migration wave, when guest workers were recruited (19603). At the same time, emigration virtually stopped. The recruitment of guest workers came to a halt after the oil crisis of 1973, migration by then had shified towards another ‘colonial stream’, this time from Surinam. When this migration 110 a/—_5~. '1 I slorrc. Nahc rneasn In and (it held hadtn unnnt eiidet nlent ofind (ChOe ln assum TCCOg: ICSCar Slluctl that The, slowed down, family reunification and an increase in refugee migration made that the Netherlands continued to have a positive net migration. With all ‘waves’, restrictive measures were searched for (Amersfoort and Penninx 1998; Schuster 1999). In the after-war period, until about the 19703, there were ‘little signs of racism’48 and discrimination in society. Surinamese nurses that came felt ‘exotic’ but did not feel this as a limitation or a stigma. There were no impediments for them, and they had more opportunities in Dutch society than in Surinamese society. Indonesian immigrants were fairly quickly absorbed in society, and there seems to be little evidence of problems. While research is lacking it would be safe to say " [that] Dutch tolerance towards blacks, colonized peoples and migrants was actually an expression of indifference towards them and of a desire to maintain a distance from them" (Choenni and Cain 1995, p. 515; Cottaar 2003). In the 19703 more research on discrimination developed. At the time, it was till assumed that immigrants were a temporary part of Dutch society, which impeded the recognition of possible (future) problems. Racism and discrimination were still research as ‘individual characteristics’, not as structural problems. In the 19803 the structural side of racism was appeared in research. The dominant (political) view was that: The groups that dominated Dutch society shared certain views: that racism and discrimination were reprehensible and unacceptable; that there was no place for racism in Dutch society; that these were more inherent to South African society or perhaps the Dutch Centrum Partij, and extreme right-wing political party; that these people who were constantly raising the issue of racism were simply not aware what they were talking about; and that blacks and migrants were too obsessed with race and the colour of their skin (Choenni and Cain 1995, p. 518). ’8 There is no research showing the signs 11] t__fl figunb other I dmpu groups the Tu Sunna Rea.l Th xxmr b): gOi inedti suflng dealin Stereo». for otl lamon Items Reubs. Th lithis l dlSCrir rel'ea] The contemporary situation is strongly influenced by the two murders of public figures: racist attacks and xenophobic discourse are now part of the mainstream. As in. other European countries, the prejudice and racism cannot be attributed to the 'one drop of blood' U.S. ideology and it is not necessarily aimed towards the 'darkest groups'. As such, attitudes towards Surinamese are not as hostile as attitudes towards the Turks and especially the Moroccans. Ethnic hierarchy research shows that Surinamese (and Antilleans) were preferred ‘others’ in the 19803 (Hagendoom 1995; Rea, Wrench and Ouali 1999). This does not mean that the Surinamese were whole-heartedly welcomed in Dutch society. As soon as the migration increased in the early 19703, public opinion, helped by government and media representations, became negative. The government actively tried to find legal measures to restrict migration, which according to some resulted in Surinamese independence. In the media stereotypes about Surinamese drug-use and dealing quickly ensured a negative context of reception for the Surinamese. The stereotyping of the Surinamese in the late 19703 and early 19803 had consequences for other aspects of life, discrimination in the labor market and in the housing market (among other fields) was high (Bovenkerk 1978/1979b; Bovenkerk and Breuning-van Leeuwen 1978/1979; Dagevos 2001; Dagevos 1998; Niekerk 2000a; Oostindie 1988; Reubsaet 1988; Sansone 1990; Sansone 1992). The attitudes towards Surinamese have become more positive, but it is not certain if this has had positive consequences. According to Gras and Bovenkerk (1999), discrimination in the labor market was even higher in the 19903. Discrimination was revealed in the application process, and in other aspects of labor force participation. 112 Their r "Despi dncnn applicz discrin the mi in the ‘ (Dagei have c minori these r ICSpOn F? Seems U”ion. r91 n.,\3 Their research revealed that discrimination did not differ across ethnic groups and “Despite their perceived greater social acceptability, Surinamese men encounter discrimination to the same degree as Moroccan men” (103). Forty percent of the applicants in their research experienced discrimination, and “ the rate of discrimination against ethnic minorities has doubled over the past 17 years” (106). On the other hand, there are indications that the labor market position of the Surinamese in the Netherlands is more stable, and less dependent on economic fluctuations (Dagevos 2001). A recent publication of the SCP revealed that Surinamese youth still have exorbitant unemployment rates: the rates might be lower than those of other minorities but they remain far higher than those of white Dutch youth. What part of these rates could be linked to discrimination, and to what extent other causes were responsible, was not reported”: Tolerance is not contradictory to racism. It is in the Dutch imagination that the one excludes the other. Analysis of everyday racism in the Netherlands revealed that active tolerance complements the repressive implications of cultural pluralism in a system of dominance. The overemphasis on ethnic difference misleads Blacks into thinking that the main goal of the struggle is the preservation of "ethnic identity," which in fact reduces culture to personality features. [...] Meanwhile the degree of exclusion of Blacks from the Dutch labor market continues to be among the highest in Europe (Essed 1991, p. 291). Racism or Ethnicism? In the Netherlands, ‘racism’ and ‘race’ are terms that are not often used. This seems to be a terminology that is mostly used in the United States, in the European Union, with its different history of ‘otherness’ and minorities, discussions focus ‘9 In: NRC Handelsblad, 21 January 2006 1 13 arounc ‘cultur M before such,: Oi and et Dutch the '0 has b, Vasta re:So< had n 0n th. “as ( Milne Willie The 1390p]E ‘ethnit around ethnicity. “Race is not the issue and that the problem of integration is ‘cultural’ or ‘ethnic” (Winant 2001, p. 141; (Lentin 2000; Vasta 2006). Migration to the Netherlands after the Second World War created a ‘racial’ other: before, nationality, religion, class were important demarcations in Dutch society”. As such, a change in racial definitions happened: This post-colonial shift has deeply altered not only demographic patterns, but also sociocultural ones. In the old dynamic the racial order and the imperial order were the same; the racialized ‘others’ were largely outside the borders of the mother-countries; they were the colonies, where they were wogs, coolies, and kaffirs (Winant 2001, p. 139) One of the consequences of World War II was a changed discussion around race and ethnicity, and ‘race’ as a basis for discrimination seemed to disappear from the Dutch lexicon: discrimination was based on ethnicity, on assumed cultural values of the ‘other’. Along with the changed discussion, a typical Dutch process created what has been called ‘minorisering’, translated as ‘turning into minorities’ (Rath 1999; Vasta 2006). With the integration of social classes into the welfare state, and with the re-socializing of the ‘lower social classes’ by the 19703, groups of social workers now had no more work. To find or create a new labor market for themselves, they focused on the integration of the new minorities. With these policies, the focus of the problem was on socio-cultural non-conformity in combination with low socio-economic status. Minorities were not defined on a racial basis (Rath 1999). Today this means that the discourse is around social and cultural norms and values, and as such, racism is seen as ‘ethnicism’. Even though this shift in focus, ’0 The position of Jews in the Netherlands was (and is) complicated. Hitler was able to racialize Jewish people in most EurOpean countries, and in the Netherlands, which might be an indication that their ‘ethnicity’ was not solely based on religion and culture. 114 from I discus a term hesitai as ‘rat‘ culturz least p Di. from biological to cultural, has taken place in more societies and has made way to discussions around ‘new racism’ (Alba and Nee 2003; Lentin 2000), ‘new racism’ as a term is not accepted in the Netherlands. This (theoretical) dilemma has created a hesitant position towards discrimination in Dutch society: people are not seen (again) as ‘racists’, but as ‘ethnicists’ and one way to ‘solve’ this is by integrating minorities culturally and socially. It does not focus on the receiving side of society, which is at least partially connected to a belief in Dutch tolerance. Discrimination compared The important question for this dissertation is if discrimination is stronger in the US or in the Netherlands: if there are different ‘levels’ of discrimination possible, and if the levels are different in the two cities, then discrimination might be an important alternative explanation. If the Surinamese in Amsterdam are displaying more upward economic mobility than the Puerto Ricans in New York City, some might claim this is because of less racism/discrimination. If the Puerto Ricans are doing better in New York, this could be explained by being in a different position as Afiican Americans, as a buffer-position, and having an advantage over the most deprived, over ‘the most’ discriminated. Finally, if no differences in economic mobility are found, one could argue that racism/discrimination have confounded the data, and that a research on economic mobility without these elements cannot be done. Discrimination could ‘undo’ the positive effects of an extended welfare state, or discrimination could undo the positive effects of the lack of an extended welfare state. One comparison concludes that discrimination towards both groups is similar (Grosfoguel 2003), others argue that one cannot compare discrimination in the United 115 States I and I it in a th and lar evidenl case fc black 2 conclu Th Comin‘ race is appltc; mnnr Nether Tacisn trump; the Ne darkec biOOd‘h Of an e", States and in the Netherlands, mainly due to the lack of a history of internal slavery and Jim Crow segregation (Pettigrew 1998; Rath 1999). Mollenkopf (2000), argued in a preliminary study that Surinamese did better, because “it appears that citizenship and language trump race in Amsterdam 1 . ..] the Surinamese second generation is evidently not transforming into a native-bom subordinated racial group, as may be the case for West Indians and Dominicans in New York City relative to the native-bom black and Puerto Rican populations” (p. 29). He did not elaborate much on this conclusion5 1. The work does pinpoint some important issues: racism and discrimination are coming from a different history in the two countries and to use the US. ‘version’ of race is problematic. Race in its biological, skin-color version is in very few societies applicable, and other defining characteristics are more important. Thus, one of the main problems is that when talking about racism, the bases are different. In the Netherlands, religion (Islam in contemporary society) is a more important aspect of ‘racism’ than is skin color or other phenotypical aspects. The conclusion that race is trumped by citizenship and language is only valid within a US. frame of reference. In the Netherlands race is not defined in the same way and to see Surinamese as ‘the darkest’ is only meaningful in a US. context. Difference is not defined by a drop of blood, and language does not necessarily ‘trump’ racism, knowledge of the language helps any group in economic mobility. Van Tubergen et al (2004) include two measures of discrimination: the inclusion of left-wing parties in government and social distance based on religion. In Europe, 5‘ The article was written in discussion with other scholars and more part of the opportunity than part of an extensive research program. 116 and specifically in the Netherlands, these measures are important, but for the United States, the extent to which they measure discrimination is questionable. In the Netherlands, and in Europe in general, religion seems to have been an important marker of difference through time. In textbooks of social problems in the US, religion is not mentioned as a base of social inequality, unlike social class, race and ethnicity, gender, and age (Beeghley 2005; Eitzen and Baca Zinn 2004). Similarly, research on social stratification in the Netherlands does not look at ‘race’, only at ethnicity. Generalizing research on discrimination based on either religion (the Netherlands), or based on race (US) can create problems because of different importance of these across the countries. This does not mean that there is no discrimination based on race in the Netherlands. On the contrary, there is ample research to show that discrimination, specifically for this research, discrimination on the job market, is rampant: estimates of discrimination of blacks or migrants by employers are between 10 and 70 percent (Gras and Bovenkerk 1999). Social distance scales and segregation measures give an indication, but do not sketch an accurate picture. On social distance, for example intermarriage rates and indices of dissimilarity, the situation in the United States appears to be far away from the Dutch (and other European) society: intermarriage rates are much lower, and segregation is incomparably high (Grosfoguel 2003). Does that mean that there is more racism and discrimination in the United States? Or does that mean that some of the consequences are different, the measures are not measuring the same, the groups measured are not comparable? 117 Be~ definit: the pot commi discri it group Amste religin C0mpl comp; detailt gIOUp estim, focus; C. factoi COncl dlSCII ’Cf) Besides measurements issues with discrimination, another problem lies in the definitions of 'characteristics’: the population from Suriname is highly diverse, just as the population from Puerto Rico. To label all these subpopulations under one common denominator, ignores the diversity, and the different experiences with discrimination. Almost no 'national' population can be labeled one racial or ethnic group. I examine the economic mobility of two 'national' groups, Surinamese in Amsterdam and Puerto Ricans in New York, in their full diversity, regardless of religion, ethnicity, gender, class, or race. As such, I will give a far more comprehensive picture of the economic mobility of the group. And as such, I avoid comparing the two groups on ‘racial' characteristics: and I will not have to give a detailed account of which group is more disadvantaged because of their 'race'. Both groups are disadvantaged through discrimination and prejudice, but it is very hard to estimate 'more' or 'less' cross-nationally. The diversity of the groups enables me to focus on the importance of context. Comparability and measurement issues make the inclusion of a ‘discriminatory factor’ difficult. The analysis will reveal if this omission decreases the validity of the conclusions. The fit of the model might suffer severely, which would mean that discrimination plays a role. In the results I will re-discuss these aspects. Segregation The racial/ethnic diversity of a metropolitan region also influences the residential situations of groups. The presence of large numbers of a minority population appears to lift the opportunities for its members to live with co-ethnics. For Asians and Hispanics, this enhancement does not generally appear to entail a cost, for the chances to reside in areas as affluent as those of the majority are not reduced; if anything, they increase. For African Americans, by contrast, the presence of a large 118 CCOl met lieu nor Cin Tesit SubL have Derc black population limits their ability to reside in affluent communities, reflecting the consequences of racial segregation (Alba, Logan and Stults 2003, p. 142). Segregation is one aspect of discrimination that directly affects opportunities in economic mobility in the United States. In the Netherlands, and specifically in Amsterdam, it is much harder to discuss segregation and its consequences since segregation levels are incomparable to the segregation degrees of the United States’ metropolitan areas. Surinamese in Amsterdam are the most segregated group in the Netherlands, but this does not necessarily relate to their position on the labor market, nor does this create a situation similar to the position of Puerto Ricans in New York City. In the United States, residential segregation is mainly a ‘black-white’ issue: Hispanics and Asians are segregated on levels incomparable to black-white segregation, and the two former groups ‘overcome’ segregation similar to the ways immigrants in the United States have generally overcome ethnic segregation. Economic mobility historically meant that people were able to move out of the ‘ethnic ghetto’, while this ‘ethnic ghetto’ never really was as separate as the residential isolation of African Americans”. The “chocolate city with vanilla suburbs” (Massey and Denton 1993, p.61) did affect Puerto Ricans: Puerto Ricans have an exceptional level of segregation from Anglos (Massey and Denton 1989). In Amsterdam, a ‘ghetto’ has not existed: the concentration of minorities was never as large as the spread of groups, and if certain neighborhoods contain a high percentage of minorities, the people are highly transitory, and the patterns of spatial 32 A ‘Little Italy’ or ‘Chinatown’ often had a lot of other ethnics living in (Massey and Denton 1993). 119 concentration change. Research in Amsterdam has not just looked at the racial/ethnic composition of neighborhoods and has included class-issues (Klerk and Amersfoort 1988; Lindeman et al. 2003; Musterd 2001). In some neighborhoods there is a high concentration of minorities and “The composition of the groups is still very differentiated, but from the limited opportunities of ethnic minorities in society, one can call these areas potential problem areas” (Lindeman et al. 2003), p. 39, my translation). Again, as in minority policy in the Netherlands, class issues are linked to minority status. Massey and Denton (1993) claim that for African Americans in US. metropolitan areas, it is primarily race that explains segregation, and that racial segregation in itself has consequences. If communities are segregated, changes in for example the economic structure can affect these communities in different ways GVIassey 1990; Massey and Denton 1993). Wilson ( 1987 ; 1988) combines race with class issues: he argues that while ‘race’ is an important factor, class issues in inner-cities are just as crucial. The disappearance of jobs, the joblessness, and the lack of opportunities in some areas create a desperate situation. This is a fairly recent aspect of segregation: while African American communities in the northern cities were always segregated, the degree of unemployment and lack of community structure changed severely with economic restructuring after the 19503. Thus, according to Wilson, the situation in inner-cities today, cannot be explained solely by racial segregation, because of this historical discontinuity (Wilson 1987; Wilson 1988). "Only Puerto Ricans developed underclass communities, because only they were highly segregated; and this high degree of segregation is directly attributable to the 120 fact that a large proportion of Puerto Ricans are of African origin" (Massey and Denton 1993, p. 12). If a large proportion of the Puerto Rican population is of African ancestry, this is not reflected in self-identification nor in identification by census- takers. Between 5 and 50 percent of Puerto Ricans are identified by census- interviewers as black, while self-identifications give estimates of 8 to 12 percent (Landale and Oropesa 2002; Rodriguez 1980b). Puerto Ricans were one of the first Latino groups to move to the northeastern cities, and us such, might have been segregated regardless of African ancestry. Just like every new immigrant group started out segregated, Puerto Ricans had their own colonias on the mainland. The important question is why there was not a similar movement towards other areas, as with other migrant group, for Puerto Ricans. Was it because of skin color, was it because of lack of opportunities, or was it because of the continuous migration to and from the island? There are many probable explanations, and the combination of explanations is most likely. The role of the (city) government can be another explanatory variable: in 1930 with a program to promote homeownership (Home Owners Loan Corporation), redlining started. With the loan programs neighborhoods were rated for investments, and poor (non-white inner-city) neighborhoods were largely excluded or ‘redlined’. The practice of redlining became bureaucratized and was applied it on an exceptional scale, and it served as model for other financial institutions. As a consequence, the development of neighborhoods was severely impaired: "Most startling was the case of New York City and its suburbs. Per capita FHA53 lending in Nassau County, New ’3 Federal Housing Administration 121 Yr an Arr Bij div con sho dep can in h. ygov. how there is 1h: York (i.e. suburban Long Island) was eleven times that in Kings County (Brooklyn) and sixty times that in Bronx County (the Bronx)" (Massey and Denton 1993, p. 54). In the Netherlands, the role of the (city) government is different. The city of Amsterdam had a policy of dispersing Surinamese from concentration in certain neighborhoods (Essed 1991)“. Both New York City as Amsterdam have high D percentages of public housing, the quality of housing standards is higher in Amsterdam. What would be labeled a ‘black’ neighborhood in the US, as the Bijlmermeer, is a highly diverse (minority) neighborhood in Amsterdam. And the diversity is not just based on race/ethnicity: In the area, a first pattern is high minority concentration and high lack of labor market participation. The neighborhood also shows a high minority concentration combined with a low degree of welfare dependency or unemployment. In short, minority concentration in the Bijlmermeer can be linked to both high as low unemployment: the report55 links this to differences in housing: some areas of the neighborhood are renovated, others solely have 'govemment housing’ (sociale woningbouw). Thus, for the Bijlmermeer, it is not clear how minority concentration and labor market participation are related. In Amsterdam, there is no pattern of "chocolate city with vanilla suburbs", the city center (old town) is the part where the least minorities live, and in the surrounding neighborhoods, minority concentrations do not seem to follow a spatial pattern. The Bijlmermeer is a transitional neighborhood, there is a high minority concentration, and this population 5’ A measure that can be racist in itself: to forcefiilly move certain groups in order to maintain a white majority in a neighborhood. In light of the severe consequences of segregation in the inner-cities of the United States, the measure seems opposite to racist: again, while racism is universal, the practices and outcomes differ cross-nationally. ’5 Amsterdamse gelto's, terugkerende kwestie. Verdieping Diversiteitsmonitor 2002. (Amsterdam ghettos, a returning issue. Broadened Diversity Monitor 2002) 122 is hi higl and part at 2 amo Iran: stror amo- level (of S ‘autr Rica Rica are r is highly residentially mobile. By 2002, the Bijlmenneer is under construction, the high rises of the 19703 are replaced by a variety of buildings with different heights and different sizes. The monotonous buildings with the separation of contact will partially make place for a modern city with a diversity of housing and people (Crok et al. 2002). “The Netherlands is not America: there are no ghettos here, nor is there the amount of social isolation as meant by Wilson” (Niekerk 2000a, p.161, my translation). Another difference with the United States is that social mobility is not as strongly correlated with geographic mobility (Niekerk 2000a). The index of dissimilarity captures degree to which groups are evenly spread among neighborhoods56 (Klerk and Amersfoort 1988; Massey and Denton 1993). The level of segregation for Surinamese in Amsterdam is high: the index of dissimilarity (of Surinamese and Antilleans combined) is 39, a number that compared the groups to ‘autochthons’ or ‘native Dutch’57 (Musterd, Ostendorf and Breebaart 1998)”. Puerto Rican segregation levels are far higher. The Index of Dissimilarity between Puerto Ricans and whites is 69 (Falcén 2005). Puerto Ricans in New York live segregated, and while Surinamese in Amsterdam are more segregated than other (minority) groups in the Netherlands, the ’6 ID= 100 * (ai-bi/2) where ai is members of group A in neighborhood 1', bi same for reference group B (Klerk and Amersfoort 1988. ’7 Surinamese and Antilleans are measured together. Even though these are groups that come from different countries with different histories, this is pretty common in Dutch literature. The effect on the index of segregation or on the index of dissimilarity is minor - their combined index of segregation is 35, and the index of dissimilarity between the groups is relatively low at 18 (Musterd, Ostendorf and Breebaart 1998). ’8 The indices have similarly high numbers in other cities: for example, The Hague has an index of segregation between Surinamese and autochthons of 47.3, Rotterdam of 39.4, while the numbers of minorities in those cities are lower. 123 consequences are more severe for Puerto Ricans in New York. In Amsterdam, housing policies, housing standards, public transportation options, closeness of alternative labor markets, all create a situation that indicates a lower level of exclusion. Segregation in the Netherlands does not have the consequences that segregation has in the United States. Social isolation in the Bijlmer is not comparable as to social isolation in New York’s inner-cities. Also, geographic mobility is not as related to economic mobility in the Netherlands as in the US. The structures of the cities, in relation to the housing market and in relation to segregation, are difficult to compare. The levels and meanings of segregation in the two countries, and in the two cities are too different to be compared. The fit of the model will show if the lack of a segregation-variable in the model distorts the explanatory power. 124 C6111 L'nit bloc deft Barr cem the ; Hob of n 196. natii 6. Welfare States The structure of the welfare state is a key feature in the contemporary process of social stratification: it creates and abolishes 'empty slots', it helps decide who fills them and how they are to be rewarded, it defines what is undertaken within them, and, finally, it shapes the pattern of mobility between them (Esping—Andersen 1993, p. 20). The Emergence of Welfare States In 19th century, two changes occurred that shaped the contemporary world for the centuries to follow. After the turbulences of the Napoleonic wars, Europe and the United States had entered a relatively peaceful period in which capitalism could bloom and develop. The delegitimation of traditional ways of governance lead to new definitions of society and to the development of the nation-state (Anderson 1991; Barrington-Moore 1966; Smith 1991). With the nation-state, ‘peoplehood’ became a central part of discourse and a force for legitimating the new-formed governments. In the 20th century, these developments continued and solidified (Bendix 1964; Hobsbawm 1990/1992; Marshall 1965; Smith 1991; Turner 2006). “The central fact of nation-building is the orderly exercise of a nationwide, public authority” (Bendix 1964, p. 22). While state-formation is a longer historical process, the functions of the nation-state increased from border-maintenance to providing an array of services. The nation-states of Europe and the United States became welfare states in such that the state became the provider (to a more or lesser degree) of health care, pension, unemployment, and family provisions (Esping—Anderson 1990; Flora and Heidenheimer 1981; Janoski 1998; Marshall 1965; Swan 1996). The increased functions of the state arose from a history of urbanization and industrialization, linked to internal and external political strive. Urbanization created 125 dependencies between the different populations of the city: health issues, the threat of epidemics, and crime rates were the bases for a more organized system of provisions. Police, sewer systems, paved streets, and other ‘networks’ of control and sanitation emerged, first only in the richer neighborhoods, later across town. These systems were the first collective goods provided by a government. Systems of support, for those who ‘fell upon bad times’ had existed before industrialization. The church always played an important role, and in the rural communities of Europe, charity was one way to ‘help’ the poor. City life again I changed this: the poor became more anonymous, and more difficult to include. A new system of charity and poor help, with poor houses and labor houses developed under control of the city governments. With the Industrial Revolution, ‘Friendly Societies’ emerged, societies of self-help among workers, but because membership was homogenous, all members were exposed to similar risks at similar times. The assistance was in times of need, for those who could afford to be members (which meant that the poorest were excluded). Corruption, the homogeneous compositions, and fraud eventually caused the state to step in: not just to provide for the poor and the workers, but also in fear of larger social consequences and political radicalization. This is where developments diverge: in some countries the state was able to interfere with support of the workers, in some the state interfered in coalition with the (large) employers, and in third states, both workers and employers allied with the government to set up a type of support system that is now called the welfare state (Braverman 1974; Flora and Heidenheimer 1981; Polanyi 1944; Swan 1996; Wilensky and Lebeaux 1965a). 126 has the uni of pet thi ins U11: “1 By the 19403 most countries in Western Europe and the United States had the basic provisions. The start of collective insurance was for sickness: the legal battles that determined who was responsible for accidents on the job and subsequent unemployment mainly favored employers, a situation that was avoided with this type of insurance. Insurance for pension plans followed, to eliminate the burden that people too old to work places on the sickness funds (old age was first defined within this type of insurance). Similarly, insurance for permanent disabilities and chronic diseases was separated from the first. A final development was insurance against the loss of income due to unemployment, the most difficult form of collective, mandatory insurance: the reasons for unemployment are dissimilar to the above three, and unemployment is far more often seen as a ‘problem of the individual’ (Swaan 1996; Wilensky and Lebeaux 1965a; Wilensky and Lebeaux 1965b). Welfare State Regimes QLiQLrLS While the historical patterns have similarities, welfare states also have different origins and different ideological bases. Social welfare developed as helping the ‘needy’ (Wilensky and Lebeaux 1965b), p. 3) and societies defined and developed different aspects of social welfare. Three traditions or alternative systems of stratification can be identified: these traditions were based in the coalitions between government and workers and/or employers, combined with an ideological view of what the poor ‘deserved’. The first is the poor-relief tradition, which promoted social bifurcation and maintained social stratification. 127 Espi \Vel: is bi hist. COIT tot 11V; lett The second tradition evolved to what is now seen as state-corporatism, and developed specific assistance for separate groups. In the ‘social insurance model, the state enhanced social stratification by consolidation of divisions under wage earners by providing different groups and classes with different benefits. A second objective of these policies was to promote loyalty to the rulers or the central authority. The third tradition started in labor movements, where first fraternal organizations were organized to ensure some social stability. This evolved in a universalistic approach to welfare, where equality of status is promoted (Barrington-Moore 1966; Esping- Anderson 1990; Swan 1996;Wa11erstein 1974; Wallerstein 1980)“. The most commonly used classification or typology of welfare state regimes is by Esping—Andersen (1990). In his book, he compared and contrasted eighteen western welfare states on two main principles. The first principle, the granting of social rights is based on the work of TH. Marshall (1965) who analyzed developments in the historical accumulation of citizenship rights. Political and civil rights were complemented by social rights with the development of the welfare state. The granting of social rights, basic guarantees for a minimal standard of living can create the process of de-commodification. De-commodification is a label used for the extent to which persons become independent of the labor market in maintaining their livelihood. Different welfare states have varying levels of social rights, or varying levels of de-commodification: the lower the de-commodification, the higher the ’9 There are many ‘causes’ to the development of welfare provisions. 1 focus on traditions towards the differences between welfare state institutions. For example, appeasement of the workers in fear of a (communist) revolution is an important background to the creation of welfare state institutions, but it does not explain why different states developed differently: this fear could be found in most western welfare states. 128 dep of I: labr Est by 1 anc citi ant strz in an en pr At 312 dependence on income generated through labor. De-commodification applies an array of benefits and provisions that enable people to be less dependent on selling their labor power to survive, and this measure is the first base for the classification by Esping—Anderson. The second criterion of clustering is the system of stratification that is promoted by social policy. This aspect looks at the role of the state in intervening in the market, and is linked to de-commodification in that it creates a buffer or safety-net between citizens and the market. The aspect also differentiates between the role of the state and the role of the family in providing for citizens. Combined, the system of stratification looks at independence of market and of family through state provisions. The two aspects, de-commodification levels and intervention of the state in the market (and to a lesser extent the family) determine the classification of welfare states in the typology. The Tygolggy The liberal welfare state supplies benefits through a means-tested system. The benefits are mainly available for low-income groups and are connected to strict rules and stigrnas. Liberal work-ethic norms have shaped social reform, and the state encourages the market in private welfare schemes. These structures lead to a predicted bifurcation of the class-system. Besides the United States, Canada and Australia fit in this liberal regime of the welfare state. The second cluster of states is around the corporative-statist-type. In this type, granting social rights never was an issue, but the granting of rights was related to status and class. The state took the role of the market to provide welfare and thus 129 private insurance schemes are marginal. Austria, France, Germany, and Italy fit best in this regime-type. The social-democratic welfare state is universalistic in its approach and provides basic benefits to all. The quality of the services and the amount of benefits are high and eligibility rests on being a citizen or a permanent resident: rights were extended to the middle classes. The social-democratic welfare state reaches these goals with a heavy social service burden. The Scandinavian countries have achieved the highest level of social-democratic welfare. Immigration and the Welfare State In modems welfare states this strength is more a question of the ability to supervise the labour market, than of policing the national borders (Brochmann 1998, p. 34). Geographically uneven economic development and the capitalistic division of labor created an emerging global economy in which inequalities between nations became greater (Castles and Miller 1993/1998; Fassmann and Mi'mz 1994; Hollifield 1992; Massey et al. 1993; Massey et al. 1998). With the emergence of peoplehood and the nation-state, immigration debates changed of topic. Before the nation-state, borders were not strictly defined, and a sense of belonging, or a sense of national identity was relatively undefined. Migration or international traveling was not controlled until the First World War, when a first system of regulation and control emerged in Europe (although Tsarist Russia had migratory regulations) (Hammar 1990; Hoerder and Moch 1996; Moch 1992). After the Second World War, several patterns of migration emerged, and by the 19803 international migration related to Western Europe and the United States 130 consists of a variety of origins and destinations (Borjas 1990; Massey et al. 1998; Munz 1995; Pedraza and Rumbaut 1996; Portes and Rumbaut 1996). The waves of migration since 1945 have lead to significant proportions of migrant populations in the western world, both in the US and in western Europe. As relatively homogeneous nation-states, countries in western Europe lack the historical experience of the United States in absorbing large numbers of migrants: there is a negative ‘ideology of migration’ where migrants are not seen as making a long-term contribution to society (Amersfoort 1998; Entzinger and Stijnen 1990; Joppke 1998b; Livi-Bacci 1993). Extensive research in the receiving societies has looked at the role of the state in its capacity of regulating migration. Policies and regulations regarding migration can be analyzed on two levels: the check and controls at the borders, and checks and controls within the borders. Countries try to regulate and control immigration by increasing border control, but also by decreasing benefits and rights within their borders (Brochmann 1998; Hollifield 1992; Joppke 1998a; Joppke 1998c; Massey 1999; Massey et al. 1998; Minderhoud 1998; Mullan 1998; Portes and Rumbaut 1996). One main form of control is the granting of ‘status’: different immigrant groups from different countries are either admitted or denied entry dependent on the view of the state of their ‘status’ (Booker et al. 1998; Hiebert 2002; J oppke 1998c; Massey 1999; Massey et al. 1998; Portes and Rumbaut 1996; Schuster 1999). Each status gives differential access to ‘citizenship’, the second level of government policies (Brochmann 1999; Hammar 1990). Since both groups entered their countries of destination as citizens, there were not legal restrictions on entitlements of the welfare state. Their status allowed them to 131 have full access (again, legally) to all aspects of the welfare state. The main difference between the level of access is that the Dutch welfare state has more extensive provisions for its citizens than the US. welfare state: The services Cubans received were superior to what was available for other immigrants at the time, or, for that matter, citizens or residents of the United States. They were the only ethnic group in the United States that received welfare 'European style' (Grosfoguel 2003, p. 112). Legal access does not guarantee full participation: “Membership in liberal societies involves not only legal definitions of who can or cannot be a citizen, but also questions of assimilation, ethnicity, race, and culture” (Hollifield 1992, p. 11). Welfare states (in Europe) are set up to ensure a minimal standard of living, not to promote full participation in all aspects of society. This problem of lack of access is not solved by economic minimal standards but would require intervention by the state on different levels. The nationalistic origins of the welfare state have caused debates on entitlements, not just for immigrants, but also towards ‘citizen-migrants’ as the Surinamese and the Puerto Ricans (Myrdal 1960; Winant 2001). Welfare State and Mobility One aspect of the welfare state is in the recreation of stratification and the connections between the state and the market (Esping-Anderson 1990; J anoski 1998). The larger the provisions of the welfare state, the less stratification within society, and the more income equality. A side effect of larger provisions of the welfare state is in that incentives to participate in the labor force can be less. This is turn can lead to a population that is excluded from higher level jobs, but does not feel the need or possibility to work in the lower level jobs, because the provisions they receive 132 through the welfare state. The relation between the provisions of the welfare state and migration are looked at within the context of the following question: Could welfare benefits (high) block economic mobility through the lack of labor force participation incentives? Since both groups came to the respective countries as citizens, they are entitled to full access to welfare state provisions. As such, we can examine the influence of the welfare state on economic positions. Comparisons of labor market positions and unemployment rates can generate insight on the influence of the welfare state on economic participation: when benefits are high and the range is large, the incentives to participate on the labor market will be lower than when benefits are low and the range is restricted60 (Borjas 1999a; Borjas 1999b; Jensen 1989; Mullan and Majka 1997). This would predict less labor market participation for Surinamese in Amsterdam than for Puerto Ricans in New York City, assuming all other factors constant. On the other hand, the array of services can enhance economic and occupational opportunities. The welfare state can provide certain training, it can aid in measures that guide demand from employers into the supply of workers, it can provide child care so that women can easier enter the labor market, and these are just a few of the possible examples. The labor market and the welfare state are not independent of each other: problems in the labor market can create intervention by the welfare state. On the other hand, the labor market (or actors on the labor market) can use the welfare state as a means to create different conditions. 5° Evidence from a comparison between Germany and the United States generated the following conclusion: "The more generous welfare state benefits in Germany make it less painfirl to stay out of the workforce, so that workers who return are disproportionately those who can do well in the labor market" (DiPrete and McManus, 1996, p.70). 133 Welfare States and Urban Areas In this research, I examine the labor market of two urban areas. These cities carry their specific problems, related to urbanization, globalization, and structural changes in the labor market. Differences in the economic base, the national institutional framework and urban policy have contributed to a variety of urban experiences in the 19803 (Kloostennan quoted in (Hamnet 1998, p.22). The notion of ‘the polarized city’ emerges repeatedly from the literature on the dual or segmented labor market. A shift from a bell-shaped curve (through jobs in manufacturing) to an inverse bell-curve (or from an egg-shaped distribution to an hour glass distribution) is hypothesized with the changes and restructuring from manufacturing to a service industry. The assumption is that ‘middle range’ jobs disappear while both highly paid, highly specialized, and low-waged, unskilled jobs become more common (Bohning 1998; Sassen 2000; Wilson 1996): The post-industrial societies can experience two alternative kinds of polarization. In the strong welfare states the polarization is between 'a small, but highly upgraded insider structure and a large outsider surplus population. In the other case, a large service class proletariat will constitute the pivotal source of polarization (Hamnet 1998, p. 25). In strong welfare states, as the Netherlands, a “large and growing unemployed and economically inactive group excluded from the labour force” can arise, while in weaker welfare states, as the US, “the growth of a large, low-skilled and low-paid labour force" (Hamnet 1998, p. 19) would prevail. The evidence for income polarization is strong, but occupational polarization does not show a similar trend, at least not in the Netherlands. The increase in income polarization is not mainly due to 134 occupational changes but can better be explained by changes in taxation, welfare benefits, and unemployment (Hamnet 1998; Kloosterman 1994). Some Cautions Income is often used as measures of economic mobility within a country, but international comparative economic mobility research uses occupational scales instead of income. Income is reflected in occupational scales, but not included (Borjas 1994; Chiswick 1979; Treiman 2001; Tubergen, Maas and Flap 2004). Thus, conclusions on poverty or income mobility are not part of this research. There is a caution for the interpretation of unemployment levels and assumptions on poverty: less labor market participation in the Netherlands does not automatically mean that (unemployed) Surinamese would be in a comparatively disadvantaged position compared to Puerto Ricans in the United States: unemployment benefits have been high (until recently) and instead of poverty, the main threat of (long-term) unemployment is marginalization (Kersbergen 2000). Unemployment (or early retirement or disability) benefits in the Netherlands can potentially provide a higher base-level income than full-time employment in a low-wage job in the United States (Dagevos 1998). The welfare state has direct and indirect influences on occupational positions. The direct (labor market) measures only form one aspect of welfare state policies. The link between occupational positions, incomes, and the welfare state is often seen through income redistribution (as measured by social expenditure as percentage of 135 GDP“, (Esping-Anderson 1990). Since income cannot be used as a measure in international comparative mobility research, other measures that affect economic positions are used. Education is an important factor on economic mobility: in both the Amsterdam and the New York City labor markets, diplomas have become very important for the attainment of a stable job. In today’s world ‘diploma inflation’ changed educational demands for even the lower status jobs. The welfare state has an impact on the labor market through education by providing educational structures. But the welfare state is not the only factor that has an impact on education: parents’ (and in particular father’s) education is still an important indicator of the educational attainment of their children (Blau and Duncan 1967; Lipset and Bendix 1967; Treiman 2001; Treiman and Yip 1989). Another problem with the inclusion of ‘education’ within the welfare state is that ‘public’ and ‘private’ have a different meaning in the two countries. For example, most Dutch universities are public, and have little differentiation in quality. The difference between public and private universities (or high schools) in the United States is very pronounces. The (welfare) states of each country are part of the separation between public and private spheres in education, but other factors play a role. The taxation for education is different, the testing standards are different, and while the systems are comparable on paper, the influence of state, welfare state, parents, and other factors are too complex to contain in a model assessing the influence of the welfare state on economic opportunities. Education is included in the 6' Gross Domestic Product 136 model as a group characteristic, an aspect of ‘human capital’ of the ethnic group. One of the measures that is used to estimate the influence of the welfare state in the model reflects the separation between public and private, using data on health care provisions (chapter 8). A final caution links to the government influence in the housing market. Although ‘housing’ is not directly within a labor market context of reception, it is a crucial aspect in relation to economic opportunities. Living in a segregated community or neighborhood reduces possibilities for mobility (Entzinger 2001; Massey 1990; Massey and Denton 1993; Musterd 200062. In the Netherlands, poverty is often looked at as ‘a lack of contact with the rest of society’, and is not linked to some ‘absolute’ standard as in the United States (Musterd and Ostendorf 1993; Musterd, Ostendorf and Breebaart 1998). Poverty is connected to three aspects: redistribution, social segregation, and opportunities for social mobility. Desegregation has been a deliberate Dutch policy towards Surinamese in the Netherlands, for different reasons (see chapter 5), but partially to avoid segregated neighborhoods as exist in the United States (Reubsaet and Kropman 1983). The two examples, on education and on housing, are ways that the welfare state has an indirect influence on economic opportunities that are not measured through (direct) labor market policies. These two indirect aspects are not included in this research: there are too many relationships between the different factors and variables to allow for the inclusion of specific measures. 62 Chapter 5 has a more detailed description of the consequence of segregation 137 United States and the Netherlands Compared Differences in Typoloqv 1n the typology of Esping-Andersen, the United States shows a low level of de- commodification, and a high score on liberalism, low on both conservatism and socialism. The position of the Netherlands is less clear: the score is high on de- commodification, high on socialism, but medium on both liberalism and conservatism. J anoski (1998), who recreated the scales used by Esping—Anderson, ranks the Netherlands as a social-democratic welfare state. The Netherlands has an odd development in welfare provisions because of the system of pillarization63 until the late 19503, which might make it difficult to fit the history in one of the three traditions (Kloosterman 1994). From the literature, it is clear that the provisions in the United States are far less than those in the Netherlands, but differences do not remain ‘constant’ over time: in both countries welfare programs are in constant change. The level of differences is explored in more detail, not just on state level, but also on local or city level. An illustration of the differences in redistribution comes from data of the OECD (OECD F actbook 2006), on the distribution of disposable income of households, Gini-indices64 reveal the levels of income inequality in the Netherlands and in the United States: since income is not included in the model, I use it to illustrate the ‘3 The division of society into three ‘pillars’: catholic, protestant, and ‘labor’, on which a variety of institutions were based, as schools, radio and television networks, political parties, etc. 6" The Grim-coefficient is a measure of income inequality. It is defined by the difference between the distribution of total income equality and the specific distribution in a country (Lorenzo curve). A score of 0 (zero) means perfect equality, 1 means perfect inequality. In the table, the Gini-coeflicients are displayed as percentages ("‘ 100). 138 redistributive effects of the welfare state: the Gini-coeffrcients for the Netherlands are lower at every time point, income inequality is greater in the United States. Through taxation and benefits, the Dutch welfare state has an influence on the income distribution in the Netherlands: Table 6-1: Gini-coefficients, mid-19803 to years around 2000 mid-19803 mid-19903 2000 Netherlands 23.4 25.6 25.1 United States 33.8 36.1 31.0 From: OECD Factbook 2006: Economic, Environmental and Social Statistics. Differences in Ideolgqy In the Netherlands, the government decided to design its Minority Policy in the interest of a category of immigrants who had only resided in the country for less than one generation and who, compared to the US. exhibited a striking absence of segregation. The government went as far as to extensively sponsor ethnic minority community organizations and even granted local voting rights to residents who were not Dutch citizens without any form of naturalization or registration and without these foreigners having campaigned for it. It is revealing that the Dutch expect their government to intervene in the country's economic and social life (Rath 2000, p. 10-11, my emphasis). Esping-Anderson looked at de-commodification on a general level, he did not focus on the ways de-commodification differed: in Germany, and even more in the Netherlands, an important way is through ‘work to welfare’: older populations were stimulated to leave the labor force: in the 19903, only two-thirds of men over 50 years old and 23% of men older than 60 are in the labor force. The ‘welfare sustaining exit policies’ are one dimension on which the US. and the Netherlands differ. Job security levels, strength of labor market boundaries, and employment sustaining labor market policies, with employment exit policies, are other dimensions of the welfare state’s influence on the labor market. 139 For example. 1996 was a year of welfare reform in both the Netherlands and the US. 1996, with an emphasis on ‘from welfare to work’. The effects of the policies are different: in the Netherlands, while the goal was to get ‘mothers’ back into the labor force, the process was on ‘integrated case —management’ (maatwerk): if a case- worker thought the mother had good reasons not to be actively involved in the labor force, she was not always forced to. “The focus on this “maatwerk” was not as squarely on getting people to abide by the work requirements as national policy makers might have liked” (Korteweg 2005, p. 26). In the United States on the other hand, the consequences were stricter and the percentage stay-at-home mothers65 decreased significantly. The policies seemed very similar, but the consequences were different: in the US, more single mothers were pushed into the labor force (Korteweg 2003; Korteweg 2005): When it came to their clients, caseworkers I in the Netherlands] saw single mothers first and foremost as responsible almost by virtue of their motherhood, where in the Burnett County [in the US] such 3:3)ponsibility was tied to paid work, not parenting (Korteweg 2005 , p. The ‘fit’ of the states in the scheme of Esping-Anderson might be different today, differences are still large enough to treat the US as a different type of welfare state than the Netherlands (DiPrete et al. 1997; DiPrete and McManus 1996; Kloosterman 1994; Korteweg 2003; Korteweg 2005). Emphasis on Work Job security levels in the US. regime are lower: the US. labor market is highly flexible with ‘fairly good’ opportunities to enter the labor market and a high level of 6’ The policy applied only to stay-at-home mothers dependent on welfare. 140 labor market participation. This is at the cost of low paid jobs and job insecurity. In European welfare regimes job security is more stable, and poverty levels are far lower, but this also had costs. The labor markets are inflexible and levels of unemployment and early retirement are higher (DiPrete 2002; DiPrete et al. 1997; DiPrete and McManus 1996; Kersbergen 2000). The US. labor market is very sensitive to structural change, the Dutch welfare state adjusts for structural changes and people’s life courses are less disrupted during structural changes in the economy labor market. Again, this has consequences for mobility levels (up- and downward). Overall, the United States comes closest to a ‘individual mobility regime’: the position of the individual in the labor market and individual resources determine outcomes (DiPrete 2002; DiPrete et al. 1997; DiPrete and McManus 1996; Kersbergen 2000; Veenman and Roelandt 1994): The Netherlands is at the other extreme from the United States in the lack of individual-level sensitivity to structural change. Our results suggest that labor force adjustment in the Netherlands has been accomplished largely through accession and employment exit (DiPrete et al. 1997, p. 351). Regulation of Work In the Dutch [welfare] state, the influence of the government on the local and ethnic context of reception is much larger than in the US. [welfare] state. Dutch policies towards immigrant integration (or incorporation) are far more extensive than US. policies, and have not just affected the legal and economic domain, but also factors of culture: there have been active policies for the maintenance of immigrant cultures. Similarly, policies for economic integration were (and are) more extensive than in the market guided welfare state of the US. (Entzinger 2001; Razin 1993). In 141 order to analyze and separate these aspects of state or welfare state governance, and their impact on economic mobility of immigrants, one has to be careful to only look at government regulation at the ‘government context of reception’. Not all government measures should fall under the label of ‘welfare state’ provisions, and not all measures are to be measured only at a ‘state level’. Finally, wages and job conditions are less regulated by the state/ government in the United States than in the Netherlands, and as such, employers can ‘use’ the different groups in the city against each other, which would create more intergroup- competition. In Amsterdam, this is less possible, for legal migrants, because of the more standardized wage and job requirements. Conclusions The labour market plays a decisive (through not exclusive) role in the integration of immigrants. Absorption of the immigrant population in the regular labour force also relieves the public budget and may contribute to economic grth (Brochmann 1998, p. 34). The type of welfare state has an important influence on the labor market. The labor market in turn has a major role in the (economic) integration of migrants. The next chapter compares the ethnic group and labor market contexts of reception. The role of the government in its welfare provisions, as a form of internal regulation, is hypothesized to be the key explanatory variable for Opportunities and economic mobility. The ‘welfare state’ as a complex of benefits and regulations has effects on (migrant) economic opportunity and mobility. Benefits, in the form of ‘de- commodifying’ measures as welfare, pensions, disability, are just one aspect of the welfare state. Labor market regulations are a second aspect of how the (welfare) state 142 can be a steering and redistributive mechanism for the economic opportunities of its citizens. These regulations can be of a direct influence on the position of minorities, as with affinnative action, but labor market regulations in general also affect minority opportunities, just as they affect opportunities for the general population. A clear example can be found in the ways wage-negotiations are set up: in the Netherlands (union) contracts automatically are applied to the whole branch of labor and not only to certain parts as in the United States. Part of these (general) regulations set the environment in which ethnic communities can develop their own means of subsistence. Ethnic entrepreneurship cannot develop in every nation-state on the same level, because of state regulations (Faist 1995; Lewin-Epstein and Semyonov 1986; Razin 1993; Wilson 1980). And the state creates jobs: for some groups, in the US. most notable for African Americans and women, government jobs (federal, state, local) have been a vehicle for upward mobility. Government jobs have been more available for minorities than the private sector, and they provide more security and better pay (Reimers and Chemick 1991). In the Netherlands government jobs provided a similar route to economic mobility, (Dagevos 1998; Vermeulen and Penninx 1994). Finally, the welfare state has an indirect effect on economic achievements through education and housing. The Dutch welfare state is different from the welfare state in the United States. These differences are on multiple levels: there are different ideological backgrounds, there are differences in redistribution and de-commodification, and there are differences in the emphasis on the dependency on work. Part of these differences are 143 tested in the model, to investigate how they affect the economic position of the groups in each country. 144 7. Group and Context Comparisons To analyze the impact of differences of welfare state provisions and benefits on economic mobility, the two other factors, “ethnic group context of reception” and “labor market context of reception” should ideally be a similar as possible. In this chapter, the similarities of these two factors are described. Differences between the groups and the contexts are carefully researched, and if these differences have shown to be significant in earlier research, they are included in the model. Position in the 19803 From the position in the 19803, a very similar picture emerges. The groups were in a very disadvantages position in both cities, had large unemployment rates, an estimated high percentage of female-headed households, high dependency on welfare, and high poverty rates. The importance of this comparison is that both groups seem to have very similar starting points, as compared to their position in the 19903 and in 2000 (2002). Puerto Ricans were not able to improve their (socio)economic position from the 19503 to the 19803, a pattern that is uncommon in the migration literature: “Island- born Puerto Rican men apparently never catch up, unless they come with no education” (Reimers 1985, p. 41). The relationship between 'length of stay' and economic position did not hold for Puerto Ricans in the early 19803, and instead of overcoming barriers and becoming to par within 15-20 years, they actually started in a position closer to the general population, and then were pushed into a position far worse. Changes in the economic structure and on the labor market of New York City 145 have had a major impact on the economic position of Puerto Ricans. Their general position was actually better in earlier periods of time. The Puerto Ricans position is very much linked to the economics of society and there seems to be little 'grip’ of the group on their own destiny within the city's economy (Sturz 1985). Surinamese immigrants in Amsterdam, on the other hand, arrived about 5 to 10 years prior to 1983. The surveys by Reubsaet et al. were at an early stage of their migration, and the positions Surinamese were in, is not as exceptional as the position of Puerto Ricans. For both groups, the changes in positions since the 19803 is examined, based on the similarities of their position in that time. From theory, certain group characteristics are important for economic mobility: length of stay, time of migration, type of migration, general education level, starting position in the labor market, average age, racial composition, location of origin (rural- urban), gender composition, family composition. These aspects are very comparable for the two groups, based on the information from that era. Another cluster of factors that has an impact on group economic mobility are contextual aspects. The structure of the cities, in regards to minority composition, labor market structure, government employment opportunities, characteristics that were strikingly similar between Amsterdam and New York in the 19803: both cities experienced severe economic restructuring, both cities had a large percentage of minorities, and of those minorities, the two groups were fairly similar in numbers. This chapter systematically analyses these aspects for the following decades, comparing the two cities. 146 The Cities Compared The city as a context of reception, and especially the labor market composition of the city, both in supply and demand, are important in shaping economic mobility. In order to examine the importance of the welfare state for group mobility, the labor market context of reception need to be as similar as possible. In the first comparison, I look at different aspects of the structure of the cities. These aspects are identified through theories on economic mobility. First, the labor force composition of the city is important: immigrant mobility theories (ethnic succession; ethnic queue; and skills mismatch) predict that areas with higher rates of migration and higher proportions of immigrants have more labor market competition. Second, the general labor market is discussed: mobility theory indicated that through economic restructuring, labor market conditions have changed and opportunities have altered. A final word is on the housing situation for the two groups in the cities. Migrants in the Cities New York and Amsterdam have long histories of receiving immigrants. Amsterdam has been and is a city of immigrants. In the 17‘h and 18‘11 century, migration to the Netherlands, and to the cities of the Netherlands, gave a significant influx of people (Beer 1998; Comelis 1990; Heelsum 1997; Klerk and Amersfoort 1988; Penninx 2005). Similarly, New York City’s history as an ‘immigration gateway’ and place of settlement is well-documented (Binder and Reimers 1995; Fainstein 1998; Foner 2003; Zhou 2002). 147 Both cities have an amazing diversity (Table 7-1) in their populations and although the New York population is much larger, the groups are similar in size compared to the total city“. Amsterdam has about 1/10‘h of the New York population. In 1999, the Surinamese population formed 9.8 percent of the population of Amsterdam. In 2000, the Census recorded 789,172 Puerto Ricans in New York City, with a total number of inhabitants of eight million — which would mean 9.9 percent of the population. ‘Foreign born’ in New York accounted for 36 percent of the population in 2000 (1990: 28.4 percent), in Amsterdam the percentage of ‘ethnic minorities’ was a little under 50 percent in 2002 (26 percent in 1990)(Kasinitz, Mollenkopf and Waters 2003; Mollenkopf 2000; 0+3 2002)“. Theories of immigrant competition thus would have similar outcomes based on these numbers. The cities have high rates of immigrants and ethnic minorities. The origins of these groups differ in each city, but socioeconomic characteristics are similar. Amsterdam might have a higher level of job queue-ing and competition than New York City, but the percentages are not completely comparable. The percentage of immigrants in the city is included in the model to estimate the factor ‘Labor Market Context of Reception’. 66 I could compare New York City to the ‘Randstad ’ to increase the general population size of the Dutch part of this research. The increase in population would not make up for the much more severe increase in difficulties of comparability of the regions. 67 Both numbers do not reflect the total immigrants: ethnic minorities in the Netherlands are only certain groups, ‘foreign born’ does not include Puerto Ricans. But both numbers do give a good impression of the diversity of both cities. 148 Table 7-1: Immigrants in the Cities New York City Amsterdam City size 8 million 800.000 Percentage ‘imrnigrants’ 1990: 28.4 percent 1990: 26 percent 2000: 36 percent 2002: 50 percent Percentage specific grog; 2000: 9.9 percent 1999: 9.8 percent From: Kasinitz, 2003; O+S, 2002; Mollenkopf, 2000. Economies of the gities Amsterdam can be considered a ‘world city’: a city connected to larger international economic systems that experience mass immigration (Rath 2002a). This is comparable to the ‘global cities’ (Sassen, 1988; 1995; 2000), cities that are: command points in the organization of the world economy; key locations and marketplaces for the leading industries of the current period - finance and specialized services for firms; and major sites of production for these industries, including the production of innovations in these industries (Sassen 2000, p. 4). While Amsterdam might be a ‘global city’ on a smaller scale, it is not the definition as much as the consequences of these new locations that matter. Both Amsterdam and New York City have seen changes in their local economy linked to their (new) position in the globalized economy. A main consequence lies in changes in the labor market. First, specialized services, involve an array of jobs, from low- to high paying: the specialists (high- income gentrification) need cleaners. This polarization in wages is increased by the decline in manufacturing. While it was assumed that the base of the city needed to be manufacturing, changes in the global economy have proven that assumption to be incorrect: the shift of manufacturing to low-wage nations, increased the centrality of global cities. The coordination of the production and distribution process is done in global cities, while the production itself is done elsewhere (BOhning 1998; Hamnet 149 1998; Sassen 1988; Sassen 1995b; Sassen 2000): "In 1950 manufacturing accounted for 42.8 percent of the city's 3.47 million jobs; in 1997 it accounted for only 15.8 percent of the city's 3.41 million jobs"(Smith 2001, p.2, on New York City). The decline of manufacturing has had consequences in itself. Manufacturing facilitated unionizing and wage-setting, which, with social benefits, became the model for other sectors. Also, consumption and production ‘went hand in hand’: the products were partially consumed by the households of the producers. Thus, it was in the interests of the employers to pay wages that enabled workers to buy the products (DiPrete and McManus 1996; Hamnet 1998; Raes 2000; Sassen 2000). For both cities, economic restructuring occurred since the 19703, and the labor market context of reception was affected by this economic restructuring. "Particularly, employment in manufacturing professions has decreased, as did employment in the manufacturing industry" (Veenman and Roelandt 1994, p.35, on the Netherlands). As global or world cities, Amsterdam and New York share characteristics of these changes. In New York, the percentage of the population working in manufacturing was about one-third in the 19503. In 1990, that percentage had decreased to 11.5 percent. In contrast, the percentages employed in the service sector went from 14 percent in 1950 to over 40 percent (43.1) for males and 63.6 percent for the female labor force of the city (Rose 1994; Sassen-Koob 1985). In the Netherlands, the shift from manufacturing to services was not as impressive: in the period 1960-1967, 40.6 percent of the jobs were in manufacturing, between 1.980 and 1988 this had decreased to 28.3 percent. The service sector had a 150 high percentage of the jobs in the 19603: 51 percent. This increased to 66.8 percent in the 1980-1988 period (Amersfoort and Cortie 1996). These are numbers for the country as a whole, in Amsterdam the industry sector lost 5200 job in the same period, while the service industry gained 33400 jobs (Kloosterman 1991). Even more indicative of the situation in Amsterdam are the percentages for 2002: in 2002, 6 percent of the population worked in the industrial sector, 44 percent was employed in services68 (0+8 2002). Job grth rates are included in the model, and some caution is needed. There are aspects of ‘job growth rate’ that are problematic. First, the grth rate does not look at the types of j obs that are created and that disappear. Second, the percentages reflect a measurement at a point in time, and they do not account for long-term changes. Third, and this is an issue that cannot be addressed by looking at one labor market, one city, in each country: data on workers outside of the city are not measured. Thus, the opportunities of commuters are not measured accurately. Overall, it is an estimate of economic growth that affects the labor market opportunities, but it is not a perfect measure of these opportunities. Table 7-2: Job growth percentages New York City69 Amsterdam70 19905 -3.4 +2 20008 +3.8 +1.08 From: http://socds.huduser.org/CPBSE/totalsodb; Kloosterman, 1991; 0+8, 2002. 6" Hotels and restaurants, financial institution, business services, other services. 69 Percentage of the change in jobs in the city 1991-1992 and 1999-2000. 70 Employment growth in the core city 1984-1988, recomputed for 1987-1988, see appendix 2 (Kloosterman 1991); employment growth 2001-2002, computed for the city, see appendix 2 (0+8 2002) 151 The results will reveal (at least partially) some of the problems: I discuss the types of jobs the groups occupy, which (partially) reflects the structure of the labor market. U-Turn in waggs The "Great U-Tum" in weekly wages had its turning point in the US. 1973 (chapter 3). The economic slow-down or down-tum after the oil-crisis of 1973 had its consequences for every group in the labor force. But the consequences were greater for some groups: "all minority groups, and especially Latinos, have experienced an even greater reversal in real and relative income levels" (Hinojosa-Oj eda, Camoy and Daley 1991, p. 26). In the Netherlands, inequality has risen in a similar way, but at a later time point and at a slower rate: while inequalities have increased in the US. since the early 19703, this process started later in the Netherlands: in the 1967-1992 period, there was first a decline in inequality, while it reversed to an increase (in the 19803). Deindustrialization (or the decline of manufacturing) is one of the major reasons for the U-Turn, and is linked to, as discussed above, 1) union membership 2) wage- setting coordination, and, as described in chapter three, 3) the de-commodification of labor by the welfare state (Alderson and Nielsen 2003; Korzeniewicz, Moran and Stach 2003). The consequences of deindustrialization are related to the welfare state: different types of welfare states had different measures to offer relieve to the (most severe) consequences of the down-tum (DiPrete et al. 1997). Other factors for the increase in inequality in incomes are increased north-south trade, rising foreign investment, increased migration, increased female labor force participation (Alderson and Nielsen 2003; Hinojosa—Ojeda, Camoy and Daley 1991). Thus, since the 19703, there are changes in levels of inequality: in the US, inequality 152 steadily rose since 1973, in the Netherlands, there was a decrease in inequality in income in the 19703, but inequalities increased after the 19803. The welfare state is a modifier in this process, with the link wage setting principles and the link to levels of de-commodification. These aspects affect the opportunities for economic mobility for society as a whole, and for minority groups living in those societies. In general, income levels are not included in international economic mobility research. It is not included as an indicator of the labor market context of reception in this research either. The general wage-level, minimum wages, and measures of inequality, are part of the discussion of the welfare state. l__gbor Mgrket Participatiognd Unemployment The general labor market participation and especially the unemployment rates affect the opportunities for immigrants. Similarly to the general increase in inequalities in wage-levels, general unemployment levels affect the opportunities for minorities. Unemployed in Amsterdam is defined as: people between the ages 15-64, registered as unemployed that a) do not have a job b) want to work for 20 hours or more c) are available to work. This excludes people that do not register (anymore), but also people that are ‘not available’: people with (physical and mental) health problems, women that decide to stay at home to raise a family, elderly and early retired. Unemployment numbers in New York City are about half of the percentages in Amsterdam. These unemployment rates are considered in the model (labor market context of reception): 153 Table 7—3: Unemployment levels and labor market participation New York Amsterdam?1 Unemployment 1990: 8.3% 1991: 14.6% 2000: 5.7% 2002: 12% of beroepsbevolkitggn Total labor market participation 1990: 3570000 1995: 64% 2000: 3650000" 2002: 73% Male labor force participation 1990: 64.2% 1995: 71% 2002: 79% Female labor market 1990: 49.2% 1995: 56% participation 2002: 67% Subsidized jobs 2002: 3.5% of beroepsbevolking *total non-farm employment, from: www.nyc.ggv/html/dcp/pdflpub/socind03ch2.pdf. Sources: O+S 1991; 0+8 2002; Rivera-Batiz 2005; Rose 1994. Government Measures Labor markets are shaped by government regulations. The most obvious measures for this dissertation are linked to affirrnative or positive action. The Dutch and US. government, and local governments have implemented policies targeting the position of minorities. As such, the labor market and opportunities on the labor market are influenced by the state. While it could be argued that these aspects are part of the labor market, I have included them as part of the welfare state. The (confirmatory) factor analysis gives a justification for this choice. Housing Surinamese live concentrated in the Bijlmermeer (or: Amsterdam Zuidoost- Southeast), an area that is part of Amsterdam but not considered the center. They are the largest ethnic minority of the city. They are mainly concentrated in Bijlmer Center 7‘ The definition of unemployed applies to those registered. For 2002, the number of unemployment comes from those registered at CW1: Center for Work and Income (Centrum voor Werk en Inkomen) Labor market participation is from 1995: interestingly, the Yearbook of Amsterdam from 1991 had a chapter on unemployment but not one on labor market participation. 7‘ Beroepsbevolla’ng: population capable of working between ages 15-65. 154 and Bijlmer Oost, where the percentages of Surinamese are 47 and 38 percent. Over 38 percent of the Surinamese in Amsterdam live in Amsterdam-Southeast (Musterd 2001; Musterd, Ostendorf and Breebaart 1998; O+S 2002). New York City has the largest amount of public housing in the United States”. Puerto Ricans live concentrated several neighborhoods or boroughs in New York City. The highest concentration of Puerto Ricans live in the Bronx, forty percent of the Puerto Ricans in New York live there, followed by Brooklyn with 27 percent. Manhattan and Queens have comparable percentages of Puerto Rican population (15.2 and 13.8 percent respectively), and Staten Island, the fifth borough of the city, has a small percentage of Puerto Ricans (3.5 percent)(Rivera-Batiz 2005). In the 19903, the Puerto Rican population of the city decreased for the first time, which meant that for the first time, they were not the largest Latino group in the city (around 37 percent of all Latinos)(Falcon 2005). In Amsterdam, there is a large share of public housing with strong bureaucratic and political control over the allocation process: Therefore we should caution against direct comparisons with settlement patterns in the USA, where success in the housing market presupposes success in a more general sense, and in the labour market in particular. Nevertheless we do not want to underrate the relative success of Surinamese in the Amsterdam housing market (Klerk and Amersfoort 1988, p. 160). For both cities, housing conditions, and housing prices are shaped by government measures. The relevance of housing is twofold: first, segregation affects economic opportunities, second, rent controls and subsidies affect standard of living. Neither of these characteristics are included in the model, but the conclusions and discussion are 73 http: www.clphaorg/page.cfm?pagelD=490 155 linked to these aspects. While the situation in both cities in housing is different, there are similarities that needed to be mentioned. Comparative Conclusions of the Cities Looking at the two cities, the situation of the Surinamese in Amsterdam more difficult: there is a larger percentage of immigrants in the city in 2002, which would lead to more competition and ethnic queueing. General unemployment levels are about twice as high as in New York City, and the job growth percentage is lower. Three measures or indicators are composing the labor market context of reception: the rates of job growth in each city, unemployment levels, and percentages of immigrants. While these do not encompass all aspects of the labor market context of reception, they do compose a good illustration of the opportunities in each city. The Groups Compared The comparison of the groups, as with the comparison of the cities, reveals similarities: these similarities are sought in order to keep the ‘ethnic group context of reception’ as constant as possible. From theory, certain aspects are distilled. First, from migration theory, the type of migration is important, as is the composition of the group linked to the type of migration. Second, human capital characteristics of the groups are examined. These characteristics have proven to be of significance in economic mobility. Finally, the composition of the ethnic groups, and their resources are discussed: besides human capital characteristics, ethnic group characteristics are important in shaping the ethnic group context of reception. 156 Migration Type Migration is linked to the intertwined history of colonial relations. As discussed before, both groups came from regions that were dominated by the country of destination. The link between the US. and Puerto Rico might better be labeled as ‘coloniality ’: "the continuity of colonial forms of domination after the end of colonial administrations produced by colonial cultures and structures in the modern/colonial/capitalist world—system" (Grosfoguel 2004, p.4). There are differences in the ‘colonial’ relation between Puerto Rico and the United States and the former-colonial relation between Surinam and the Netherlands. Puerto Rico was colonized by the Spanish, and institutional arrangements were shaped after Spanish institutions. Language, religion, cultural elements, these all were infiltrated by Spanish domination. In Surinam, the English were the first colonizers, but for a very short period (1650-1667). In 1667 the Dutch took over, and shaped the colony in a Dutch system. Institutions as education and language were affected, and are still influenced today (although there is a discussion today to change the language of education into English). That does not mean that there are no overlaps with colonial regimes: there is a relationship of dependence and of domination. “Other European countries, particularly France, but to a lesser extent the Netherlands, had colonial policies of ‘direct’ rather than ‘indirect’ rule and assimilated their Caribbean colonies as part of the nation state, rather than accepting the inevitability of ‘independence” (Cross and Johnson 1988, p. 90). Colonial relationships and colonial migrations are not ‘the same’. Even within a country as the 157 Netherlands, the history of migration of the different former-colonial groups gave differences: the context of reception changed over time and the colonial relationship with the countries was variable (Amersfoort and Niekerk 2006; Schuster 1999). The crucial importance of the colonial relationship is the long history of connections: migration is triggered by these connections, especially when citizenship is allotted to the (former) colonialized peoples (Cross and Entzinger 1988; Cross and Johnson 1988; Essed 1991; Grosfoguel 2003; Grosfoguel 2004; Model and Lapido 1996; Sansone 1990; Sansone 1992). Another consequence of the colonial relationship is that a variety of people migrate. Unlike for example guest worker migration or refugees, a cross-section of society migrated in a colonial type of migration. This is not different for the two groups studied. Finally, the groups came to the destination countries as citizens. The commonalities of migration of the two groups can be found in l) colonial movements 2) towards metropolitan areas 3) long history of connections 4) in the 19703 excluded from the labor market 5) being kept in a low position by racial ideology (Essed 1991; Grosfoguel 2004). mersity of the Groups The sense of peoplehood shared by members of a racial-ethnic group identifies their culture as a distinct experience and way of life, commonly expressed in a collective name. The answer to the question "who are we?" may change, as the peoplehood changes, because of major events such as migration or inclusion in another society by war, occupation, and dominance (Hernandez 1996, p. 128). The Surinamese population in Amsterdam and the Puerto Rican population in New York City are diverse groups. Migrations started as ‘elite migration’, for 158 educational purposes for Surinamese, for general purposed for Puerto Ricans in New York. Migration for both groups increased with the introduction of cheaper (air) fares, and a broader representation of society was able to migrate (Choenni 1997; Grosfoguel 2003). Ethnically, the Surinamese population in Amsterdam roughly consists of 60 percent Creole and 30 percent Hindustani. Other Surinamese groups are, and the list is not exhaustive, Javanese, Chinese, Jewish, Lebanese, and ‘Dutch' ( ‘Hollanders ’) (Choenni 1997; Heelsum 1997). Puerto Ricans in New York are a heterogeneous group. Racial backgrounds are varied: while "The majority of Puerto Rican woman in the US. and Puerto Rico are classified as white on the birth certificate" (Landale and Oropesa 2002, p. 241), self- identifications are more complicated. F orty-two percent of Hispanics rejected the census categories on race, and select ‘other’ because of the fluidity of racial classifications on the island. Eight percent of women self-identified as ‘black’: if it was required to be classified into US. categories: 5 percent were identified as black, 32 percent as ‘not white, not black’, 34 percent as ‘possibly white’ and 29 percent as ‘unquestionably white’ (Landale and Oropesa 2002; Rodriguez 1980b). Finally, there was usually no identification with 'Latino' or 'Hispanic' but with the national group (Grosfoguel 2004; Hernandez 1996). Overall, the racial diversity is complex. Both groups are difficult to ‘fit’ into categorizations of race and ethnicity. This is the main reason for studying the groups based on their (country of) origin. It will give a more complete picture of the opportunities of the groups, regardless of race. 159 For Surinamese in Amsterdam, diversity also exists by level of education, length of stay, and rural-urban origin (Choenni 1997; Dagevos 1998; Gras and Bovenkerk 1999; Heelsum 1997). Similarly, Puerto Ricans in New York have very diverse backgrounds and represent the population of the island by social class, urban-rural origin, and education levels (DeFreitas 1985; Haslip—Viera and Baver 1996; Ortiz 1986). These similarities justify a comparison between the groups on economic mobility: as described, these are all factors that affect economic mobility of groups. It is impossible to include all these aspects in the model, first, because of differences in meaning (as described in chapters 4), second because of data set limitations, and third because of the severe complexity and redundancy this would bring". Education levels While ‘education level’ is a measure that is used in international comparative studies, there are some issues with this variable. First, countries have different educational systems. Second, within the United States, there are differences in schools, colleges and universities. ‘High school completed’ does not mean that a similar level of knowledge has been reached across high schools. While these latter issues are important, and should be taken into consideration, the data sets do not provide for this specific information. To compare educational categories from the Netherlands to US. categories, I have relied on a classification used by DiPrete et al. (1997). For the Netherlands, the categories that are included are less than primary education and primary education; 7“ For example, the data sets for the Netherlands do not include the proposed variable ‘percentage urban-rural’. 160 secondary education, lower level (low vocational training, low general education); secondary education, upper level (intermediate vocational training, high general education); tertiary education, lower level (vocational colleges); tertiary education, upper level (university). These categories capture both the Surinamese and the Dutch educational systems. For the US, some different categories are applied: less than high school completion; high school completed; some post secondary education; bachelor’s degree or higher (DiPrete et al. 1997): Table 74: Education categories New York Amsterdam Less than high school Less than primary education Primary education High school completed Secondary education, lower level Some post secondary education Secondary education, upper level Tertiary education, lower level Bachelor’s degree or higher Tertiary education, upper level From: DiPrete et al., 1997. The relative position of each group within each society on education is not altered by these differences in categorizing. For example, in both the US. and the Netherlands, educational demands have increased over time. Diploma inflation is important in both countries, but it is relative to the demands before. Thus, while ‘college’ might be a requirement of a certain level in the US, ‘college’ has its own meaning in the Netherlands. The groups are compared within each society, and not cross-societal: I do not look at the differences in education per se, I look at how educational levels affect economic mobility within each society. In 1980, around 64 percent of Surinamese in the Netherlands had primary education or less. Of Puerto Ricans in New York, 62 percent had not completed high 161 school. While 25 percent of Puerto Ricans in the city had a high school diploma, 23 percent of Surinamese in the Netherlands had an equivalent educational level. Nine percent of Surinamese had a higher secondary education completed, and nine percent of Puerto Ricans had ‘some college’. College or more for Puerto Ricans was slightly higher than upper level education for Surinamese 4 and 2 percent (Cruz 2005; Reubsaet, Kropman and Mulier 1982). The literature indicates that both groups have improved their educational position (Falcon 2005 ; Rivera-Batiz 1991; Rivera-Batiz 2005; Vermeulen and Penninx 1994). Self-Employment Self-employment and entrepreneurship is an alternative route to economic mobility (Kloosterman, Leun and Rath 1998; Kloosterman, Leun and Rath 1999; Light, Bhachu and Karageorgis 1993; Portes and Jensen 1989; Portes and Zhou 1996; Raes et al. 2002; Rath 2002b; Waldinger 1989). It is part of the group context of reception, and as such can be a group characteristic that can skew the data on economic mobility: if one group displays higher self-employment and has this route to mobility, this would affect the hypothesized importance of the influence of the welfare state. Entrepreneurship or self-employment does not play an important part in their labor market position: about 3.5 percent of Surinamese in Amsterdam is entrepreneur (Boissevain and Grotenbreg 1988; Choenni 1997; Reubsaet 1988). The Puerto Rican community has a viable entrepreneurial sector in the early stages of migration. Through urban renewal and urban development, this sector has virtually disappeared (Glazer and Moynihan 1963; Torres and Rodriguez 1991): 162 Indeed the Puerto Rican economic and political experience in the United States has largely been shaped by resource allocations in New York City. Given the lack of a viable entrepreneurial base and a continuing history of discrimination among private employers, the progress of a large segment of the Puerto Rican population has been influenced by local governmental decisions about resources for education, welfare, and jobs (Stafford 1991, p. 159). Because of the lack of significant numbers of self-employment of both groups, this variable is not included in the measurement of the ‘ethnic group context of reception’. Female Headed Households The economic position of a household is dependent on several aspects: Mobility in household conditions is determined by changes in the labor market conditions of all household adults, by changes in household composition, and by state policies that mitigate the socioeconomic consequences of events that potentially alter a household’s standard of living (DiPrete 2002, p. 267). While DiPrete discusses ‘life conditions’, which includes more aspects than occupational mobility, the article gives an excellent overview of measurement issues in mobility. ‘Income producing assets’ are broader than the (monetary) gains from occupations. ‘Secondary income streams’ increase the overall standard of living, but are also generally less stable than the income of the breadwinner. The standard of living is not only dependent on the streams of income, but also on the number of dependents in the household. Finally, social welfare state programs are not (always) linked to the income of the breadwinner (DiPrete 2002). Overall, certain factors determine the ‘class status’ of a household, and female-headedness is one of those factors. 163 In the Netherlands, non-married partnerships are institutionalized and recognized as partnerships equivalent to marriage, there is less stigma than in the United States on couples that live together, have children, and do not aim to get married. From the literature, it seems that Surinamese mothers that choose not to marry are problematized and labeled as 'single mothers' living in 'female headed households' (Essed 1991). This process is even stronger for Puerto Rican single mothers (as most single mothers) in the U875. The main problem of single parents is that their odds for living in poverty are greater than for households with two parents (in the labor force). This is especially true for female-headed households. Labor market participation and occupational standings are related to household compositions. The positions of female-headed households are different in both countries, and these differences seem to boil down (again) to differences in the welfare state. Besides a greater incidence of poverty in the United States due to less extensive benefits, female-headed households are in a more precarious position because of differences in regulations. 7’ Something that might not have been the case in the early 19803: the demographic overview of Surinamese in the Netherlands includes many characteristic, but ‘female headed household’ is not of them. The table on household composition included unmarried, married, widowed, and divorced. From the table, the percentage of Surinamese (men and women in the Netherlands) that is unmarried between the ages 25 and 34 is higher than the percentage of the total population: 18.8 for the total and 31.7 for the Surinamese. Overal, in the Surinamese population 42.4% reports to be unmarried. Divorce rates also differ, 2.7 percent for the total population, and 9.6 percent for the Surinamese. These number indicate that there are a higher percentage of female-headed households for Surinamese in the Netherlands (Reubsaet, Kropman, and Mulier. 1982). For Puerto Ricans in New York City, the percentage of female-headed households was 33.2 percent (Cruz, 2005). 164 Time of Migration “Bovenkerk has once noticed that the tragedy was that it [the migration] was at the wrong moment in time” (Niekerk 2000a, p.21, my translation). Surinamese entered the Netherlands during an economic downturn: the mid-19703 was a time of crisis in the economy, unemployment was high, and housing was scarce. The peak of Puerto Rican migration to New York City was 20 years earlier, in a period of relative prosperous economic circumstances and little migration to the United States. The situation changed drastically with economic restructuring and renewed competition through increased migration. Their situation, despite the earlier start, was comparable to that of the Surinamese in the 19803. The model is run with a time-factor included. Out-migra_tion Some of the literature mentions that successful Puerto Ricans move out of New York City, but recent studies indicate that “For Puerto Ricans, suburbanization and return migration were not just the purview of a growing middle class, but also of those at the bottom of the socioeconomic ladder” (Falcon 2005, p. 158). The va y ven phenomenon is an important aspect that can have an impact on the mobility of the group: the composition of the group differs with this continuing cycle of migration. But, on the other hand, if the group composition (in terms of education, age, and other human capital characteristics) does not change, the cyclical phenomenon only has its relevance in ‘years of migration’: since there is a constant inflow of new immigrants, the time of entry is not easily determined. This could be one of the reasons that Puerto Ricans in New York City, as a group, have not been as successful in reaching ‘wage parity’ as other immigrant groups. It is difficult to determine the extent of this factor 165 since there are no borders crossed in the migration process, and records of outrnigration are lacking. Language We have found little support for the idea that exposure to the destination language before migration enhances labor market opportunities. One exception was the position of male migrants who had moved from a country in which the destination language was official [...] Language exposure did not affect the employment chances of females, however, not did it have any effect at all on labor market participation (Tubergen, Maas and Flap 2004, p. 718). Surinamese came to the Netherlands with knowledge of Dutch. The majority speaks Dutch at home and is fluent, although there are signs that there are some language problems (Essed 1991; Gras and Bovenkerk 1999; Koot and Uniken Venema 1988). Puerto Ricans entered the United States from a Spanish-speaking country: "Nevertheless, birth on the island is analogous to being born abroad because these women are likely to have been socialized in a Hispanic environment, including having received their education in the Spanish language" (Tienda and Guhleman 1985, p. 246). On the other hand, a large majority of Puerto Ricans in New York indicate they speak English, approximately 90 percent are bilingual ‘in varying degree’ (Pérez y Gonzalez 2000; ‘Rivera-Batiz 1991). This can be an important difference in their economic incorporation. Research results on the importance knowledge of the language of the receiving country is not clear-cut towards a beneficial advantage: “However, fluency in Dutch and familiarity with the school system, seem to have no great positive effect”(Veenman 2001, p. 235; Alba, Logan and Stults 2003; Mullan and Majka 1997). On the other hand, there are results that indicate that a lack of English does affect labor market opportunities 166 (Reimers 1985; Torres and Torre 1991). These latter results might be strongly correlated with (other) educational characteristics, as education levels increase, knowledge of the language increases, and the labor market position improves (Tienda and Guhleman 1985). Finally, an oppositional effect on language is displayed at employment levels in the federal government: speaking another language at home increased employment levels for the state (Reimers and Chemick 1991). In general, over 50 percent of Puerto Ricans indicate to speak English very well, while another 25 percent has ‘moderate’ knowledge of English (Rose 1994, for 1990). Over 90 percent of Surinamese have fluent knowledge of Dutch (Driessen 2004). For both groups, the younger, the better the knowledge of the language. With some caution, language is not included in the model. The variables make the categorization difficult: in the 1991 SPVA are there are a number of variables to estimate the knowledge of Dutch of the respondents. Not all of these variables are the same in the 2002 SPVA. The Census uses yet other categories. Besides these categorizing problems, the effects of language on economic mobility are not clear. And finally, the language barrier can be different in each country: speaking Spanish in the US. is not a similar limitation as speaking Sranang Tongo in the Netherlands. Comparative Conclusions of the Groups The comparisons of the groups give a picture of advantage for Puerto Ricans: Puerto Ricans have been in the United States for a longer period of time than Surinamese have been in the Netherlands. On the other two variables, education and female-headed households, the differences seem minimal. 167 The next chapter discusses how the measures on the ethnic group contexts of reception were obtained from each data set. The first measure is on time of arrival: since Puerto Rican migration to the mainland started earlier than Surinamese migration to the Netherlands, this aspect needs to be included to see its effects. Second, the average education is included in the model: as education is an important factor determining future occupation, the average education of the groups has to be compared. And third, household composition is included as a dummy-variable: female-headed households are experiencing more difficulties in economic attainments. The larger the percentage of female-headed households in a group, the lower economic mobility will be. Welfare States The above discussion stipulated and explained the measures for the ethnic group characteristics and the (city) labor market conditions. The similarities in the groups and in the labor market context of reception, leave one additional factor: the government context of reception. From the above two contexts of reception, Puerto Ricans were in a theoretically slightly better situation: labor market conditions seem more favorable, and the earlier time of entry creates a better ethnic group context of reception. On the other hand, there is the probability of a larger percentage of female- headed households in the Puerto Rican community than in the Surinamese. The main differences in the welfare states of the Netherlands and the United States lie in the level of de-commodification (independence of income from labor), measured through ‘public social expenditures as a percentage of GDP, Table 7-5), as explained in chapter 6. The larger the public social expenditures, the broader the 168 welfare state provisions in a country. The second dimension is ‘redistribution’, which is measured through taxation measures: taxes on the average production worker as a percentage of the total labor cost. Taxation is one important way to redistribute income, and to restratify society. From Table 7-5 and Table 7-6 several points can be derived. The percentage of social expenditures went slightly down in the Netherlands, from 1990 to 2001. The reverse happened in the United States, expenditures increased slightly. Overall, the public social expenditures as a percentage of the GDP stayed higher in the Netherlands. Taxes on the average production worker (as a percentage of labor costs) went down in both countries over the time period, but with a larger percentage in the Netherlands. Overall, again, the numbers were far higher in the Netherlands. Table 7-5: Public social expenditure as a percentage of GDP 1990 1991 2000 2001 Netherlands 27.65 27.73 21.77 21.75 United States 13.37 14.42 14.19 14.73 From: OECD F actbook 2006 Table 7-6: Taxes on the average production worker as a percentage of labor cost 1991 2000 2002 Netherlands 46.5 45. 1 42.5 United States 31.3 30.8 29.7 From: OECD F actbook 2005 Finally, an indicator of the amount of ‘public and private’ responsibilities within a welfare state is included: the total and public expenditure on health (with public as a percentage of total expenditures, as explained in Appendix 2). In the Netherlands, the percentage of public health care expenditures, is far larger than in the United States. The percentage did go up for the US, while it decreased in the Netherlands: 169 Table 7-7: Percentage public expenditures as a percentage of total expenditures (see appendix 2 for calculations), on health care expenditures. 1990 2000 2003 Netherlands 67 63 62 United States 40 44 44 These measures give an estimate of the differences in welfare state expenditures in the United States and in the Netherlands. As such, combined with the measures on the ethnic group characteristics and the labor market conditions, a model can be estimated where the major influences on the labor market position of the groups can be assessed. The following chapter explains the model, the measurements, the methods, and the data sets in detail. 170 8. Methods and Data From the descriptions of chapter 7, the situation in the two cities indicates a more difficult or competitive situation for Surinamese in Amsterdam: the percentages of unemployment are higher, there is a larger percentage of immigrants in the city, and job grth percentages are lower. The group comparisons also give a picture of advantage for Puerto Ricans: they have been in New York City for a longer period of time. Education levels are comparable. The percentage of female-headed households might affect this advantage negatively (but data on the percentage of female-headed households are not available for the Surinamese in Amsterdam). This chapter discusses the data sets and how the measures on the ethnic group contexts of reception were obtained from each data set. The second part of the chapter covers the methodological approach. Finally, the different steps of the analysis are described and the model is developed in detail with the variables included. Datasets The time-analysis is done over the years 1991 and 2002 for the SPVA in the Netherlands, 1990 and 2000 for the Census. These years are selected for various reasons. First, for the Dutch data sets, the years closest possible to the census years were selected. Second, while Surinamese migration increased drastically in the mid- 19703, the first SPVA was held in 1988. The extensive research by Ruebsaet et al. (1982/1983) is not compatible with the data from the SPVA, and overall, the 1990 Census was the earliest possible comparative year. In order to keep the Dutch period close to the 10 years of the Census, 1991 was selected over 1988. 171 The SPVA in the Netherlands The survey Sociale Positie en Voorzieningengebruik Allochtonen (SPVA — Social Position and Utilization of Provisions by Allochtones) is a longitudinal survey conducted by the Erasmus University Rotterdam- ISEO (lnstituut voor Sociaal Economisch 0nderzoek: Institute for Social Economic Research) in 1991 and by the same Institute combined with the SCP (Sociaal C ultureel Planbureau: Social Cultural Planning Agency) in 2002. The first SPVA was conducted in 1988, and it is held among the four largest minority groups in the Netherlands: Turks, Moroccans, Surinamese, and Antilleans. It is a comprehensive, in-depth survey on education, labor market position, social-cultural aspects, and interactions with autochthones. The questionnaire includes a detailed overview of (national) backgrounds, age, marital status, household composition, education in and outside of the Netherlands, diplomas, reasons for lack of diplomas, labor market history, participation, type of employment, unemployment and its reasons and time period, income, housing, knowledge of language, and other aspects. The goals of the SPVA are threefold: the first goal is to understand the socioeconomic and social-cultural position of the four largest minority groups, the second goal is to map differences in the position of the different groups, and contrast them to autochthones. Thirdly, the longitudinal aspect is used to understand changes over time (this latter goal was not specified at the SPVA 1991). Data Specification The SPVA has two parts in the survey: the heads of households and family members. The questionnaire for the family members is not as detailed as the series of questions submitted to the heads of household. Only the latter surveys can be used in 172 this dissertation, the first series (2002) does not include information on employment or education. A second step in the data specification was to select the Surinamese population in Amsterdam“, and two filters were imposed on the data, since the SPVA covers the main minority groups in the main cities. The third step involved the selection of Surinamese population of working age. The International Labour Office (ILO)77 defines the working age population between the ages of 15 and 64. This restriction excludes those parts of the population that are too young or too old to be part of the labor force. The total respondents is reduced by these necessary restrictions, to 196 in 1991 and to n=179 in 200278. The US. Census - IPUMS The United States Census is held every 10 years, and accounts for a large part of the population living in the United States. Because of the enormous size of the data set, I have used the Integrated Public Use Microdata Series (IPUMS) of the University of Minnesota, where Census data are compacted into accessible quantities. The IPUMS consists of samples of the American population drawn from the Census and uniform codes are assigned over different Census-years and samples”. With the sampling of IPUMS, it is possible to pre-select a certain city (New York City = 461) and a certain group (Puerto Ricans = 2 on ‘hispang’; Hispanic Origin). For the next filter, on age, I adjusted the lower limit. After looking over the data, it appeared that the inclusion of 15-1 8 year old would distort the data. These 7" “etngroep” was limited to value ‘3’, Surinamese; Amsterdam: gemeente =1 77 From: www.ilo.org/public/english/employment/strat/kilm/ 7" The SPVA-surveys that I use are held among ‘heads of households’ which means that there are no young people included. The youngest age is 19. 9 From: http://www.ipums.org. 173 respondents are still mainly in school, and since the Surinamese data do not contain people younger than the age of 19, I decided to make the minimum age same as the minimum age of the Surinamese respondents. With these filters implemented, the final count of each dataset is a population of 19800 in the Census 1990 and in 2000, the number of total respondents is 18858. Variable Selection — Erthnic GrougContext of Reception From the surveys among heads of households, group characteristics are assembled. Ethnicity, residence, and age are first selections, and to measure the factor ‘ethnic group context of reception’, the additional variables are ‘time of arrival’, ‘average starting education’, and ‘household composition’. Time of arrival is measured in years: in both the 1990 Census and the 2000 Census ‘years in the United States’ is the most comprehensive measure. This variable is measured in categories, and the intervals of the SPVA91 and SPVA02 are adapted to these categories (using ‘verblkl’ for 1991 and ‘verblijf’ for 2002: see appendix 2 for details). As with ‘time of migration’, the Census has been used as the basis for the measurement of ‘average education’: the 1990 and 2000 Census have the same categories (although some were collapsed), the SPVA was not as consistent over the years. The re-categorization of the SPVA labels is explained in appendix 2. Household composition (gezinssamenstelling in the SPVA) is measured directly in both years. The values 1 and 7 represent female-headed households, and these are re-categorized into dummy variables where female-headed households are labeled 1. 174 For the Census, a more complex transformation was needed (again, detailed descriptions of the data transformations can be found in appendix 2). Additional Measures In chapter 7 the measures for the labor market context of reception and the indicators for the welfare state context of reception are described. In the previous section of this chapter, the measures for the ethnic group context of reception are detailed (in appendix 2, an overview of all the measures is given). The final measure that needs specification is the dependent variable, economic mobility. Economic Mobility The measurement of the dependent variable, economic mobility, is through the occupational classification by TLC (ISCO-88), as described in chapter 4. The use of ISCO-88 is fairly straightforward for international comparisons, the categorization is based on an international standard. There are ten major groups, 28 sub-major groups, 116 minor groups, and 390 unit groups, organized around skill-level and job requirements. The occupations of the data sets are recoded according to this classification, in a step-wise manner to translate national codes as accurately as possible to the ten major ISCO-88 groups (for details, see appendix 2). The first step was to recode the classifications of the SPVA-91 into the same occupational categories as the 2002 SPVA. These classifications follow the Standaard Beroepenclassificatie 1992 (Standard Occupational Classification or SBC92), a list of over a thousand job descriptions. The SBC92 includes occupations as ‘wooden shoe finisher’ (26201 afichrijver klompen), specific descriptions of types 175 of teachers (as ballroom or folk-dancing, 43301), and is too detailed to use for a comparison. The subcategories, as ‘lower technical’ (272), ‘scientific mathematical and science professions’ (851) offer a base for recategorization that has a lot in common with the ISCO-major groups, with some exceptions. The occupational categorizations are enumerated in table 8.1, following the categories of ISCO88. This table also includes the codes of the Census. In 1990 a different coding scheme was used than in 2000, but the categories largely stayed the same. The Census codes overlap with the ISC088 in many details, and only a few adaptations had to be made. The main difference between the Census and the ISC088 is in the description of ‘elementary occupations’, a category that was fairly straightforward to compose from the descriptions of the jobs. The final steps in the recategorization were to exclude farming occupations and military occupations: these categories are not represented in the data. For easier interpretation, I also reversed the categories: in the ISC088, the higher number, the lower the occupational classification. The final categories (labeled: Occu_final) mean that the lower the category is numbered, the lower the occupational status or classification. 176 Table 8-1: Occupational categorizations ISCO (1) (2) (4) (5) (6) (7) (8) (9) (0) Census 1990 003 through 022 023-037; 043-199 203-259 303-391 263-274; 283-290; 413-447; 456-469 473-499 503-699; 866-874 703-865 275-278; 357; 403- 408 433-435; 443; 448-455 875-890 (military: 980-983) 992 unemployed Census 2000 1 through 33 34-186; 200-326 190-196; 286; 290; 296; 330-373; 375- 395; 900-904 500-593 374; 400-410; 430- 471; 475-494 600-613 620-762; 780-785; 874-890 770-775; 790-873; 912-960 411-416; 420-425; 472; 495-496;961- 975 (military: 980-983) 992 unemployed SBC 1992* 93, 98 55 56 62 63 69 73 75 76 77 78 79 82 83 84 85 88 89 9192 95 96 97 492 493 53 58 59 64 65 66 67 68 7172 74 86 87 314 315 484 513 514 515 2123 25 29 316 317 34 37 4142 43 471 473 485 494 495 516 517 518 54 57 24 44 262 263 268 271 460 461 462 463 466 467 468 472 521 264 265 266 267 272 28 464 465 482 483 11261 Occu_ final From: http://www.ilo.org/public/en,glLs_h/bureau/statfrsco/iscoSS/publ3.htm: by skill level (the higher the number, the more complicated skills are required), the major groups are in parentheses). " All numbers have 5 digits, the numbers displayed are the first two or three of the total. The recoded categories are organized in the following way: category 1 refers to ‘elementary occupations’ as messengers, street cleaners, and other street services. 177 Category 2 includes all ‘Plant and Machine Operators and Assemblers’. Category 3 is used to describe jobs in ‘Craft and Related Trade’. The next category includes all service related work, as sales (retail) occupations, travel services, the food (restaurant, hotel) industry, and a variety of other service occupations. Category 5 is less inclusive and represents ‘administrative support occupations’ or ‘clerks’. Categories 6, 7, and 8, are the ‘higher occupations’. Category six are ‘Technicians and Associate Professions’, category 7 includes ‘Professionals’. The last category is composed of ‘Legislators, Senior Officials and Managers’. Methodology The analysis of the data, of the two groups in the two time periods is complex. Most research on economic mobility looks at the changes in position without including different aspects that can be the cause for these changes. I look at the differential influences of the welfare states, controlling for other important variables. This complexity has created an analysis in two steps. The first step looks at the time periods to see which group was more mobile or experienced more change in occupational position. In the second step the three contexts per group per time period are analyzed, to check if indeed the welfare state was the important factor in explaining the economic position of the group. ComparinLTime Periods Both the SPVA and the Census do not link the data to individuals: “Each of these samples is independent; it is not possible to trace individuals from one census year to the next” (www.1pums.org). As such, it is not possible to look at individual economic 178 position changes over time. The first step of the analysis looks at changes in the overall economic position of each group, using Analysis of Variance (ANOVA). This method of analysis allows for group—comparisons over time, comparing the differences in position. One problem with the group comparison is that the Puerto Rican sample sizes are far larger than the numbers of Surinamese respondents. If sample sizes are large enough, any changes will be significant: “Indeed, virtually any small true difference can be made a significant one if the researcher gathers enough observations” (Weiss 2006, p. 129). In order to avoid this problem, I use two-way ANOVA: the two-way ANOVA does not just look at the individual groups, but can combine the groups over time and use the interaction effect to compute differences in change (Agresti and Finlay 1997; Turner and Thayer 2001; Weiss 2006). Other advantages are that ANOVA is robust to deviations of the normality assumption and to a lack of homogeneity of variance (Weiss 2006). For an ANOVA, several data requirements need to be met: ANOVA is a statistical method that compares the means of groups, to analyze if these are significantly different. Thus, a first requirement is that there is a meaning to the comparison: if there is no meaningful order in the data, the comparison of means has no interpretable value. The data need to be measured at least at an ordinal level. With the dependent variable recoded into 8 categories this requirement is met. ANOVA can use both quantitative as qualitative variables as explanatory (or dependent) variables. 179 In some cases, I do report 95% confidence intervals retrieved from one-way ANOVAgO. The intervals are influenced by sample size, and thus the ranges of possible values differ for the two national groups. They do not give information about the significance of the differences, but confidence intervals are illustrative for the possible value of the mean of a group (Blalock 1960; Weiss 2006). One of the limitations of ANOVA is that it is sensitive to multicollinearity: it is sensitive to combined influences of different measures. It is not a suitable method of analysis to look at the influences of the different contexts of reception because each context is measured by multiple variables. In the second step the importance of each context is evaluated and estimated using Factor Analysis. The first step in the analysis answers the first two questions that are specified in the introduction: 1. Is the occupational position of Surinamese better than the occupational position of Puerto Ricans? This question assumes that the mean occupational category of Surinamese in 2002 is higher (on Occu-final) than the mean category of Puerto Ricans. 2. Has the occupational position of Surinamese improved more than the occupational position of Puerto Ricans? Is the interaction between year and group different for Puerto Ricans than for Surinamese? The interaction term shows the differences of the combination of group and time labels. If the interaction term is significant, this means that there has been differential changes for each group over time. 3" In appendix 3 I provide more information on the differences between one-way and two-way ANOVA and other methodological issues in the steps of the analysis. 180 Comparing Contexts ANOVA is not a good method to look at multiple indicators for a same concept. The model outlined in . Figure 1-1 Contexts of Reception (Introduction) specifies three contexts of reception: the ethnic group, the labor market, and the government context of reception. These contexts are measured by multiple variables, and by using factor analysis, I have included several aspects that have an impact on each context. To measure these contexts using multiple variables, few statistical techniques are available. Factor analysis is a method to reduce the number of variables to a lesser number of factors that contain variables that measure a similar concept. From factor analysis, one can retrieve ‘regression factor scores’, or combined scores of the variables that can be used in regression. A second method that uses factors is structural equation modeling (SEM). Factor analysis assumes that the observed, or measured, variables underlie a concept, the factor. There is a system in the observed variables, and factors can be used to reduce the variables to some common component. From the theory, three factors are assumed: the ethnic group context of reception, the labor market context of reception, and the welfare state context of reception. This theoretical model is displayed in, including the measures. The theoretical model assumes that these observed variables are linked to the connected factor. This needs to be confirmed through a confirmatory factor analysis (CF A). In a CFA, the number of factors is predetermined (while in an exploratory factor analysis, the computer program assesses how many factors can be used). If the CFA confirms the factors, which is 181 analyzed through the ‘variance explained’ and through factor loadings, then the factors can be used in a regression. Factor analysis assumes normal, interval variables (at the measurement level). There are solutions for deviations of normality thought the extraction method. For normal data the most common extraction method is Maximum Likelihood (ML), a method that is sensitive to deviations of normality, such as categorical or binary data. Two of the observed variables in the data sets of this research are non-normal: female-headed household is a binary (with only two values) variable, while ‘maximum education’ is a categorical variable. These deviations of normality indicate that another extraction method is wanted, such as WLS (weighted least squares), WLSMV (robust weighted least squares), and ULS (unweighted least squares). I evaluate these extraction methods in chapter 9, the analysis and results. Another (theoretical) assumption in confirmatory factor analysis is on the relationship between the factors: are there correlations? Oblique rotations81 allow the factors to be correlated, an assumption that is highly likely (if not certain) for the different contexts of reception (Brown 2006; Kim and Mueller 1978; Thompson 2004) The main questions that is answered by this second step are the third and fourth questions of the introduction: 1. Which ‘context of reception’ has most impact on the occupational position of each group in each city over time? 8‘ Rotations are used to ‘clean the data’ and while they do not improve the fit, different rotations are available to estimate the coefficients. 182 This question is exploratory: how are the paths from the different contexts of reception to the occupational position of the groups? The second question is more specific: 2. Is the welfare state (as the government context of reception) the most significant influence on their economic position? It is assumed that the welfare state has a more significant influence on the position of the Surinamese because of the most extensive provisions in the Netherlands. The Model In the model, the ethnic group context of reception is measured through a combination of 1) length of stay, 2) education level, and 3) household composition, defined as percentage of female-headed households. The labor market context of reception is measured on a city level, using percentages of unemployment and immigrants in the city. The third measure for this context of reception is the percentage of job growth in each time period. The third context of reception, the welfare state, is measured on a national level: the first observed variable are ‘social expenditures as a percentage of the Gross Domestic Product. The second variable is the taxes on the average production worker as a percentage of total labor cost, the third is the percentage of public health expenditures as a percentage of total health expenditures (see chapter 7 for more detail). The welfare state can influence education, and welfare state measures affect conditions in the labor market. In the initial model, I have excluded the relationship between the welfare state and education. This relationship would be deterministic, the 183 welfare state influences education but not vice versa. This would imply a different model with education (as part of the ethnic group context of reception) dependent or endogenous to the welfare state. The relationship between the welfare state and the labor market is modeled as a correlation (without causal assumptions): the welfare state can have an impact on the labor market, and vice versa. Certain changes in the labor market, certain demands from employers or employees can change welfare state policies. On the other hand, the welfare state can interfere in the labor market because of other reasons than (direct) demands. 184 Figure 8-1: The Complete Model82 Length of Stay V Education Level / / Female Headed Household Ethnic Group Percent Immigrants V- Labor Market p Percent Unemployment k Occupational Position / Percent Job Growth Social Expenditures \ Tax on Labor r Welfare State 1 Public Health Expenditures 8‘ For simplicity, I have left the measurement errors out of this picture. Observed or measured variables are represented as a square, factors or latent variables are depicted as a circle, as is common in CF A. 185 9. The Analysis and Results This chapter first gives an overview of the frequencies and descriptives of each group for both years of analysis. The second part looks at changes on the level of the ethnic group and their significance levels over time through two-way ANOVA tests. Finally, occupational categories per group and across time and place are analyzed. The ethnic group characteristics that are used in the analysis are length of stay, education levels, and household composition. The first step in the analyses is an overview of these characteristics per group and per yearg". Descriptive Statistics on the Group Characteristics The first part of this chapter looks solely at the ethnic group context of reception. It was assumed, from the literature of the 19803, that the group compositions were very similar. The first step is to analyze if this is the case in 1990 and 2000. The Surinamese population of Amsterdam has more women than men. In 1991, a little over 55 percent of the respondents are women, while men compose around 44 percent of the Surinamese respondents in Amsterdam. In 2002, 60 percent of the respondents are women, a little under 40 percent are men. In the Puerto Rican population of New York, there were more women in the data set than men (55 to 45), the ratio was stable across the different census years and very similar to the ratio of men and women among Surinamese in Amsterdam. 8" The years of analysis differ slightly per group: when I discuss ‘ 1990’, I should include ‘ 1991’ for the Surinamese, for legibility I have not done so: when 1990 or 2000 is mentioned in respect to both groups, I mean 1991 and 2002 for the Surinamese. 186 Length of Stay The variable ‘length of stay’ reflects the change of time-period used in this dissertation: in 1991 the percentages of Surinamese in categories 3 and 4 (indication a length of stay between 10 and 20 years) were still fairly high, in 2002 the main concentration is in category 5: over 20 years. Considering that the main migration period for Surinamese to the Netherlands was in the 19703, the change in the length of stay is not remarkable. Figure 9-1: Length of stay, Surinamese in Amsterdam, 1991 and 2002 1991: p”... ._. 70-‘ 60”1 50-1 40— Percent 304 20—4 10— I 1 l l l 1 0-5 years 6—10 years 11-15 16-20 21+ years born in years years Years In country 187 2002: Percent 0“ I r 1 r 1 I 0-5 years 6-10 years 11-15 16-20 21+ years bomin years years Years In country Even though their stay is shorter than that of Puerto Ricans in New York City, the length of the time period is such that theories of immigrant economic adaptation would assume parity with the general population (see chapter 4). Thus, differences in length of stay are not as important as might have seemed at first glance. The length of stay in the United States for Puerto Ricans follows a similar pattern: there are less respondents in the lower categories in 2000, the categories with a shorter time in the United States. The number of new people in the city is remarkably low and does not indicate a ‘va y ven ' phenomenon of circular migration for Puerto Ricans in the city. 188 Figure 9-2: Years in the United States in 1990 and 2000 Percent Percent 1990: 0‘L I 1 JL 1 I r 1 1 1 O-Syears 6-10years 11-15 16—20 21+years bomin years years Years In country 0 l 1 1 r 1 J 1 1 1 1 0-5years 6-10years 11-15 16-20 21+years bomin years years Years In country 189 Education Levels In Figure 9-3 the increase of education levels of Surinamese in Amsterdam over the ten (eleven) year period is displayed: 190 Figure 9-3: Education levels in 1991 and 2002: 1991 BO—f 50* 40‘ 30—1 Percent 20- 10— 1 1 ‘ 1 1.00 2.00 3.00 4.00 highest level of education Percent l l l 1.00 2.00 3.00 4.00 highest level of education 191 Education levels for Puerto Ricans in New York City seem to have changed to a much lesser extent: there is an increase in levels, but the graphs do not display a change as dramatic as that of Surinamese: 192 Figure 9-4: Education levels of Puerto Ricans in New York City 1990 60-4 l l l l 1.00 2.00 3.00 4.00 highest level of education 2000 l i l 1.00 2.00 3.00 4.00 highest level of education 193 The changes in education for each group prompt an important question: how was the increase in education level for Surinamese from 1991 to 2002 achieved? The educational position of Surinamese in 1991 had already improved since the 1980, when educational levels of Surinamese and Puerto Ricans were still very similar (as described in chapter 7). From table 9.1, the distribution of the percentages on educational levels in 1990 and 1991 is already different, but not to the extent of the differences in 2000 (2002). Table 9-1: Educational levels in the early 19903 and in 2000 (2002), in percentages Educational Level (as coded Surinamese Puerto Surinamese Puerto in chapter 7) 1991 Ricans 1990 2002 Ricans 2000 1.00 35.7 49.3 14.0 40.4 2.00 30.7 25.7 29.2 26.9 3.00 23.5 19.1 36.8 23.9 4.00 9.9 6.0 19.9 8.8 Total 100.0 100 100.0 100.0 Female Headed Households The third ‘ethnic group context of reception’ variable, ‘female headed households’ shows a picture that was not expected from the literature analyses. Both in 1990 (1991) as in 2000 (2002), the percentage of female-headed households was higher among the Surinamese in Amsterdam than for Puerto Ricans in New York City. In 1990 31.2 percent of Puerto Rican households were female headed, and 41.8 percent of Surinamese households. By 2000 both percentages had increased slightly: 32.6 percent for Puerto Rican families in New York City and 44.1 percent for Surinamese. 194 Table 9-2: Percentages Female-Headed Households 1990 (1991) 2000 (2002) Surinamese 41.8 44.] Puerto Rican 31.2 32.6 While the percentage of female-headed households among Puerto Ricans in New York is in line with the percentage in 1980, I did not expect the number of Surinamese female-headed households to be higher. There was no number available in the 19803, as such, my expectations were not based on empirical data. A possible reason for my distorted expectational pattern might be the emphasis that is placed in the literature in the United States on the issue of female-headed households. This created a notion that the ‘problem’ is more severe under Puerto Rican families. Conclusions on the Numbers The frequency tables show a number of things across the groups. First, the male and female ratios are very similar. Second, the length of stay of each group in each country in 2000 (2002) is sufficiently long to not have an impact on their economic position in relation to migration year. Another important result is that the percentages of Puerto Ricans that are new on the mainland, are not high. It is safe to assume that the ‘va y ven ’ phenomenon (chapter 3) does not apply to these respondents. Another important finding is the change in educational attainment for Surinamese in Amsterdam, as opposed to the lack of change (or the very little change) in education levels of Puerto Ricans in New York City. This would give reason to investigate an alternative model, where the influence of the welfare state on education is examined. 195 Finally, the household compositions of both groups give high numbers of female- headed households, the percentage for Surinamese is higher than that for Puerto Ricans. The picture that emerged from the literature on the 19805 has changed by 1990. In the 19805, the group descriptions were very similar, with low educational levels, a high (expected) number of female-headed households, and an adaptive phase for the relatively new migrants. These characteristics have diverged by 1990 (1991), for different reasons. The length of stay had stabilized, and the majority of both groups have been at the place of research for over 20 years. The percentage of female-headed households in the Puerto Rican data is fairly stable, with a slight decrease since 1980, and a slight increase between 1990 and 2000. An exact number on the percentage of Surinamese female-headed households in 1980 is not available, but considering that 42.4 percent of the total population is unmarried (in 1983), the percentage of female- headed households must be lower. In the 1991 and 2002 this percentage has drastically increased. The most divergent result is in education levels: these had risen much faster for Surinamese since 1991. Already, in 1991, education levels were higher for Surinamese than for Puerto Ricans, and this difference seems to have increased in 2000 (2002). Step 1: Changes over Time The second part of this chapter, changes over time, looks at the changes of the characteristics of the groups between 1990 and 2000 in more detail, and evaluates the significance of these changes. This step has become necessary due to the differences 196 of the group characteristics in the data sets. The assumption that the ‘ethnic group context of reception’ has similar impact on occupational position, might not hold. Education Levels In order to assess the changes in educational levels, a common method is to compare the means through a one-way analysis of variance (ANOVA). From the data, and from Figure 9-3, the differences in education of the Surinamese over the years are highly significant (appendix 3). The problem in analysis comes with the results of the significance test for Puerto Ricans: according to the test-value, their educational increases are even more significant. The (significance) test results do not reflect the results from the graphs. To illustrate more, 95% confidence intervals give an estimate (with 95% certainty) of the value of the mean in education: Table 9-3: 95% Confidence intervals Surinamese and Puerto Ricans, over time. 95 % Confidence Interval for the Mean Lower Upper Group Year Bound Bound Puerto Ricans 1990 1.8038 1.8301 Puerto Ricans 2000 1.9969 2.0254 Surinamese 1991 1.9419 2.2248 Surinamese 2002 2.481 1 2.7704 From the table, the differences between the educational changes of Puerto Ricans and Surinamese can be read. Even though the significance test (in appendix 3) shows a higher significance level for the change in Puerto Rican educational levels, the graphs and confidence intervals tell a different story. The far larger sample sizes of the Puerto Rican data sets in both time periods affect the significance levels. 197 The difference in sample size is a problem that distorts the statistical analysis. In order to avoid this problem, I ran two-way ANOVA. In a two-way ANOVA, the testing is done in a way that sample sizes are combined and used to compare and contrast all groups. Two-way ANOVA bases its analyses on the total number of respondents, in this case, n=3 8958. It uses differences in outcomes between all four groups, and as such it controls for sample size“. The changes in the mean-education of the groups of graphed in 1F igure 9-585: Figure 9-5: Changes in mean education per group Estimated Means: highest level of education 4.00-— year of 2 study 3 —2000 E ----—w2002 g 3.00— 1991 E’ ——‘1990 a 2 '0 _ 3 2.00 8 re .E a re 11: 1.00— I I Surinamese Puerto Rican group of study The graph shows the increase in the mean level of education of each group. It also shows that the mean starting point for Surinamese was higher, and it gives a good 84 A feature that is not available in one-way ANOVA: if I wanted to test the significance of the changes in a one-way, I would have take a random sub-sample of the same size of the Surinamese samples for the Puerto Rican population. 85 I have included a different graphical display of the changes in appendix 3. 198 visual picture of the larger increase in educational level for Surinamese. In Table 9-4, the significance of the difference in changes in levels of education is confirmed (year*gIOUP)t Table 9-4: Test of differences in the changes of education levels De endent Variable: highest level of education Type III Sum Mean Source of Squares df Square F Sig.__ 3:21:99" 455.864(a) 3 151.955 161.462 .000 Intercept 6530.846 1 6530.846 6939.446 .000 year * group 455.864 3 151.955 161.462 .000 Error 36649066 38942 .941 Total 180049.000 38946 Corrected Total 37104.930 38945 Unemployment rates a R Squared = .012 (Adjusted R Squared = .012) Before I analyze the changes in occupation, I look at changes in unemployment rates. The unemployment rates might not be indicative of a shifi in occupational positions, they do illustrate a shift in participation in the labor market, and can reflect changes in the general labor market structure. As described in chapter 7, general unemployment rates in the cities declined, while the unemployment rate in Amsterdam remained higher than that of New York City. Table 9—5: General unemployment rates in each city (from: Table 7-3, chapter 7) New York Amsterdam?" Unemployment 1990: 8.3% 1991: 14.6% 2000: 5.7% 2002: 12% of beroepsbevolking 86 The Beroepsbevolking. population capable of working between ages 15-65. 199 In 1990, 31.3 percent of Puerto Ricans in New York City reported to be unemployed. This had slightly decreased to 29.3 percent in 2000. Surinamese experienced a sharper decline in unemployment rates: in 1991 the rate was at 46.9 percent, almost half the population of working age in the sample was not active in the labor market. In 2002 this number had decreased to 33.5 percent. Overall, the unemployment rates of the groups were drastically higher than those of the general population. The ANOVA reveals that there was a significant change in employment rates for Surinamese from 1991 to 2002. For Puerto Ricans in New York, the change in employment rates is significant, but again, the sample size has an enormous impact: the values of the 95% confidence intervals are very close. In general, the one-way gives little information about the changes. The two-way ANOVA illustrate that there are differences in the changes, and the difference is again significant (F of the interaction is 14.749, more information is in appendix 3). Figure 9-6 gives an overview of the changes over time: unemployment levels decreased for both groups, but as can be seen in the figure, they decreased to a larger extent for the Surinamese in Amsterdam: 200 Figure 9-6: Changes in unemployment levels Number of unemployed: 0=employed, 1=unemployed 1 .00—4 year of study on ———2000 c Q -~-—— 2002 q, 0.80-— 2 1991 T. ——-1990 .E 0.60-— E’ a 5 0.40—« '0 3 8 E 020 .3. - n in I“ 0.00- ‘ 1 1 Surinamese Puerto Rican group of study Occupational Positions Table 9-6 summarizes the occupational categories (described in chapter 8). I first report the frequencies and the changes over time per group, then I describe the differences of the changes between the groups. Table 9-6: Occupational Levels General Description of Occupation: Elementary Occupations Plant and Machine Operators and Assemblers Craft and Related Trades Workers AWN—i Service Workers and Shop and Market Sales Workers Clerks Technicians and Associate Professionals Professionals OO\IO\LII Legislators, Senior Officials and Managers 201 F reggencies in Occupational Categories The position of Puerto Ricans has improved: the percentage in professional jobs has increased from 10 percent to over 16 percent in category 7 and from a little over 5 percent in category 6 to over 10 percent. In the category of legislators, senior officials and managers, there was a very slight decrease. Categories 4 and 5 have remained fairly stable: there is a slight decrease in Puerto Rican representation in category 4, the service sector. The decrease is also seen in category 2, while categories 1 and 3 remained stable. Figure 9-7: Changes in occupational position Puerto Ricans, 1990 to 2000: Occu_flnai, 1990 30-l Percent N O 1 10.. ° 1 1 1 1 1 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 202 Occu_flnal, 2000 404 301 Percent N O 10— o._ll_._._ 1‘00 2.30 3.20 4.Too 5.100 elm 7700 elm For Surinamese, the increase in category 8 is small, there is a decrease in the percentage of Surinamese is category 7. The increase in category 6 is high, the percentage of Surinamese holding a professional occupation in 2002 has almost tripled since 1991. Category 5 (clerks) has seen an increase, and in category 4, the percentage of Surinamese remained high: over 30 percent of the Surinamese in the samples hold a service job. Categories 2 and 3 have declined, while the percentages in category 1 are stable over time. 203 Figure 9-8: Occupational changes Surinamese, 1991 to 2002 Occu_flnai, 1991 40- 30— ‘E O r 201 10—~ 0 I l 1 1 1 1 1 1 1 1 1.00 2.00 3.00 4.00 5.00 6.00 7.00 3.00 Occu_flnal, 2002 40— 30—« 8 20... E 10—« ° F f 1 n 1 1 1 00 2 00 3.00 4.00 204 Significance of the Changes Two-way ANOVA is used and the graph of figure 9. illustrates the changes over time per group. Three results come from the graph: first, the starting position in the early 19905 is higher for Surinamese. Second, the increase in position between 1990 and 2000 is larger for Puerto Ricans than for Surinamese between 1991 and 2002. Third, the positions in occupations do not have seem to change much (the mean values remain between categories 4 and 5). The first result means that relatively to the 19805, Surinamese were able to improve their position. And while the increase in position-level is larger for Puerto Ricans, the main occupational level achieved by 2000 is still (slightly) lower than the main occupational level achieved by the Surinamese in 2002. The last result, the relative lack of change is especially remarkable when compared to the changes in educational levels for the Surinamese. The improvement in occupational position is not as large as the increase in educational levels. 205 Figure 9-9: Changes in mean occupational level, per group Estimated Means: Occu_final 8.00— year of study a ———2000 g 7.00-l Hugooz o a 1991 '5 600—1 990 r: - 5.00- 2' a o E 4.00— 0 g 3.004 5 ca 2.00# 11.1 1.00—4 l W Surinamese Puerto Rican group of study From Table 9-7, the interaction effect is significant again, but the F-value, F=60.325, of the interaction is lower than that of the interaction effect in education (where the F—value is 161.462). This interaction means that the groups have experienced significantly different changes in occupational categorization, but not as significantly different as the changes in education were. Table 9-7: Test of difference in changes of occupational categories De endent Variable: Final Occupation Categories Type III Sum Source of Squares df Mean Square F Sig.___ Corrected Model 766.105(a) 3 255.368 60.325 .000 1019113891 16564.646 1 16564.646 3913.055 .000 year ' group 766.105 3 255.368 60.325 .000 Enor 114723.279 27101 4.233 Total 595017.000 27105 Corrected Total 115489.383 27104 a R Squared = .007 (Adjusted R Squared = .007) 206 For the Surinamese in the sample, there are barriers that block the educational gains from being translated into occupational gains. Changes in Education Levels and Occupational Categories In this part of the chapter, changes in occupational means are displayed per educational category. Figure 9-10 reports the results for the Surinamese. People with a bachelor’s degree or more, were able to improve their occupational position (again, as a group, this is not looking at individuals). For those that had more than high school, but less than a bachelor’s degree, occupational levels went down. Similar, for those with little education (less than a high school diploma), the occupation levels decreased. A bifurcation in opportunities seems to be taking place. In the next step, when the model is checked on the data, possible causes, as the changing labor market structure, and possible the welfare state, are examined. If the model does not reveal any possible cause, there is good reason to believe that discrimination on the labor market might play a role. 207 Figure 9-10: ‘Average’ occupation level per education level over time, Surinamese in Amsterdam Estimated Means of Occu_final, per education level, Surinamese 8.00 year of study ————2002 7.00—l ~~—— 1991 6.00— 5.00-— 4.00—1 I3“ 3.00~ / 2.00 1.0.3. Estimated Marginal Means l l l I 1.00 2.00 3.00 4.00 maximum education obtained For Puerto Ricans, this picture is different: per education level, not much occupational change is reported: 208 Figure 9-11: ‘Average’ occupation level per education level over time, Puerto Ricans in New York City Estimated Means of Occu_final per level of education, Puerto Ricans 8.00~ Census year —-2000 —~———1 990 s» >1 O O O O 1 1 5.00-J 4.00- Estimated Marginal Means 88 11 1.00— T l l l 1.00 2.00 3.00 4.00 highest level of education Changes in Educational Levels and Employment The differences in the changes in occupational position by education levels for Surinamese prompt another question: how are unemployment rates related to education levels? While general unemployment levels went down for both groups over time, there are variations across educational groups. For Surinamese, the levels of education give a large different not just in levels of unemployment, but also in changes in the levels of unemployment. The groups with education level 2 and 3 have seen the largest decrease in unemployment levels, unemployment for the groups with highest and lowest levels of education is relatively stable. 209 Figure 9-12: Unemployment levels per educational category, Surinamese Estimated Marginal Means of percentage of unemployed, per education level, Surinamese 0.70—1 year of study ————2002 0.60- -~-1 991 0.50— 0.40— 0.30—i 0.20— Estimated Marginal Means 0.10— l l l I less than high school some post BA and high school completed secondary higher highest level of education For Puerto Ricans, the trends across education levels are very similar as to their occupational levels trends. For each education level, unemployment rates are lower for Puerto Ricans in New York City than for Surinamese in Amsterdam. 210 Figure 9-13: Unemployment levels per educational group, Puerto Ricans Estimated Marginal Means of percentage of unemployed, per education level, Puerto Ricans 070— year of study -——2000 060—1 - ~ - 1990 0.50~ 0.404 0.30-9 0.204 ‘ 0.10—1 \ 0.00— Estimated Marginal Means I l l 1 less than high school some post BA and high school completed secondary higher highest level of education Step 2: The Importance of Context For the assessment of the importance of each context, a first step in this part of the analysis is to see if the variables combine into the specified contexts. The theoretical model assumes that the observed variables are linked to the connected factor and it is a necessary condition for structural equation modeling. If the ‘measurement model’ is accurate, then a structural model can be analyzed. This needs to be confirmed through a confirmatory factor analysis (CF A). In a CF A, the number of factors is predetermined (while in an exploratory factor analysis, the computer program assesses how many factors can be used). The three factors, or contexts, that were derived from theory, need statistical confirmation. 211 Another important condition of structural equation modeling is that there are enough ‘knowns’ to estimate ‘unknowns’: statistically (and mathematically), you cannot estimate a number of parameter that is larger than the number of known parameters. The number of known parameters needs to be larger than the number of free parameters. Parameter Estimation From Figure 8-1: The Complete Model, a problem emerges in the estimation of the model. With the covariance between the labor market and the welfare state context of reception (a necessary theoretical adjustment), the model has become under-identified: the number of known parameters is less than the numbers of parameters to be estimated. A (theoretical) solution would be to include another measurement variable, but in the case of this research that is not easy. One issue with additional measures is that the data set needs to increase: ironically, while in the first step the Puerto Rican data sets were too large to do one-way AN OVA, in the second step, the Surinamese data sets are ‘too small’. “In CFA, the knowns are usually the sample variances and covariances of the input indicators” (Brown 2006), p. 64): the sample variances and covariances are derived from the measures (or indicators), the addition of measures creates more elements in the input matrix. Adding measured variables causes sample size issues: there are no determined rules for the relationship between sample size and measurement model, but there are rules of thumb. In general, a sample size larger than 200 is recommended for models with ten or more measured variables. In case of the model from figure 8.1, with nine measured variables, two suggestive rules indicate that the model can be used with the 212 sample sizes of the SPVA: 1) sample size should be at least 50 more than eight times the variables in the model 2) at least 15 cases per measured variable. Since the model contains 9 measured variables, each sample size should be larger than 135 (second rule). Increases in the measured variables will give less stability of the results”. Another problem with additional measures is availability of the measures: the three ‘ethnic group context of reception’-characteristics were available for both groups over time. Other characteristics that have theoretical relevance were not as available (or would cause other problems, as described below). Factor Analysis The factor analyses reveal that while the labor market and government context of reception variables give a good percentage of variance explained, this is not the case with the ethnic group context of reception. Ideally, a composite measure as a factor should account for a large percentage of the variance explained: for the ‘ethnic group context of reception’ this is only 37 percent (Appendix 3). This initial percentage of variance explained by the three variables that should compose the ethnic group context of reception asks for further analysis. As a first step I used a different extraction method: Unweighted Least Squares (ULS), available in SPSS. Commonly, in factor analysis, Maximum Likelihood (ML) extraction is used. One of the problems of ML extraction is that it is sensitive to deviations from normality (or: not robust for non-nonnal variables) and that it cannot 87 From: www2.chass.ncsuedu/garson/pa765/structur.htm. I am using the stricter rule to be safe, there is controversy about the ‘exact’ way of determining the minimal sample size. 213 extract factors when there are binary variables”. In my data sets, both were present, and ULS is described as a possible alternative. In the literature, this method of extraction is not guaranteed to be a solution, and the problem of ceiling or floor effects can remain serious (which occur when there maximtun and minimum values to the variable, as with education)(Brown, 2006). The caution has proven to be valuable in my analysis: the ULS-extraction89 did not improve the first factor, ethnic group context of reception (Appendix 3). The aspect ‘female-headed household’ does not ‘load’ high on the factor. An exploratory factor analysis, where the number of factors is not predetermined, shows that while ‘length of stay’ and ‘education’ compose a factor, ‘female-headed household’ is a separate variable that should not be included in the factor ‘ethnic group context of reception’. To have a factor estimated by only two variables is problematic: “Measurement models in which factors are defined by only two indicators per factor can be problematic, however, so larger samples may be needed to obtain a converged and proper solution” (Anderson and Gerbing 1987, p. 14). Factor analysis does not deal very well with non-normality, and I could not confirm the three contexts of reception that were hypothesized to have an influence on the occupational positions. Without the results of the confirmatory factor analysis, I cannot run a structural equation model (since the measurement model is not 3" I have recoded the variable ‘female-headed household’, which was a Dummy-variable with only the values (1) and (0) into a percentage value, using the results of the first step of analysis. This did not improve the model, the original variable is used in the final steps. 89 There are other alternatives, these are not available in SPSS, and have similar problems. 214 appropriate), the original goal of the methods, nor can I use factors to estimate the influence of the different contexts of receptiongo. Alternative Approach? Without valid results of the CFA, there are few alternatives to estimate the influence of each context. The variables cannot be used as separate indicators because of issues of multicollinearity. Another problem is that with the lack of fit for the ‘ethnic group context of reception’-factor, the other factors cannot be estimated. Without the ethnic group variables, the two other contexts become obsolete. Not only would a model without group characteristics be incomplete, the way the two other contexts are measured gives too little variation. Each factor would be composed of only twelve different values. The most viable alternative to these statistical issues is a theoretical approach. In the conclusion I discuss how the differences in educational change and in occupational position can be understood. 9° Another way in which I tried to improve the factor loadings was by using rotations. In this case, a rotation was used that is appropriate in combination with the assumption that the factors are correlated (or oblique). Again, this did not result in improvement. 215 10. Conclusions and Discussion This research has encountered many methodological problems, which has made it impossible to draw statistical conclusions about the influence of the different contexts on economic position and economic mobility for the groups in the two cities. That does not mean that no conclusions on the relationship between economic position and the welfare state can be drawn. The conclusions can only be supported statistically in a limited sense. In the first part of the conclusions, I look at the group aspects, their occupational positions, their educational achievements, and how these can be linked to welfare state theory. In the second part I look at statistical topics. Finally, I discuss alternative ways to (try to) investigate the influence of welfare states on group positions. Conclusions on the Results There are conflicting trends in the data: on the one hand, the position of the Surinamese in Amsterdam is ‘better’, with higher education levels and a slightly higher mean occupational position. On the other hand, Puerto Ricans have been able to improve their occupational position in the studied time period to a higher degree, even though their education levels have not changed in par with the Surinamese education levels. Changes in occupational position and unemployment levels are fairly stable for Puerto Ricans on each education level, while for the Surinamese respondents in Amsterdam, the changes are related to educational level. Figure 10-1 displays the 216 changes in unemployment for each group, organized per time period and per education level. 217 Figure 10-1: Group, education, and changes in unemployment percentages at highest level of education = less than high school 0.701 group of study 9% ——Surinamese 060‘ —--~Puerto Rican 0.50~ 0.40- 0.30— 0.20—l 0.10.1 0.00— Estimated Marginal Means l l 1.00 2.00 time period of study at highest level of education = high school completed 0.70 group of study —6urinamese 0.60— — Puerto Rican 0.50— \ 0.40-4 0.30—- 0.201 0.10-1 Estimated Marginal Means 0.00 l l 1 .00 2.00 tIme period of study 218 at highest level of education - some post secondary 0.70.1 group of study 2 ——Surinamese : 0-60—1 ~~—Puerto Rican f 0.505 8 '- 0.40— E’ as 2 0.30- 3. 0.20— \o g 0.10- '" 0.00 l l 1.00 2.00 time period of study at highest level of education a BA and higher 0.70— group of study 2 _l ——6urinamese g 0'60 -———Puerto Rican g o.5o~ r: '- 0.40-l 9 E 0.30—4 3 0.205 s ._ _. a 0.10-— _ . '3 0.00— I l 1.00 2.00 tIme period of study The unemployment levels for the Surinamese in Amsterdam are higher at each educational level than those of Puerto Ricans in New York City, but they decreased 219 more over time, while unemployment levels for Puerto Ricans increased slightly (except for those with ‘some post secondary education’). The reported decrease in the unemployment of two percent for Puerto Ricans (chapter 8) is not generated by all education levels, only by those with ‘some post secondary’ education. There are several possible explanations for the higher general unemployment levels for Surinamese. The first is following assumptions of the ‘welfare state trap’: welfare states with higher benefits have higher unemployment levels because there is little enticement to work instead of receiving high benefits. The second possible explanation lies in labor market conditions: there is a higher percentage immigrants in Amsterdam than in New York City, there is more competition, and there has been less job growth in the period of time. Labor market conditions are less favorable in Amsterdam than in New York City. A final explanation can be linked to discrimination: from the literature, reported discrimination on the labor market is high in the Netherlands. The combination of discrimination, unfavorable labor market conditions, and higher benefits is most sensible as an explanation for the higher unemployment rates. Each explanation by itself is not sufficient: if the welfare trap was the main explanation, why are unemployment levels higher for all educational groups? Welfare state benefits might be ‘more lucrative’ for low-educational categories, this is not the case for the higher educational levels. It certainly does not explain why the largest decrease in unemployment for Surinamese was in the second education level, high school completed (and the third level). The welfare trap-notion similarly does not explain why the general unemployment of Puerto Ricans in New York City did not increase, but decreased with 2 percent, even though welfare state 220 provisions went up on two of the three measures in the United States. Overall, the data seem to oppose an explanation based on ‘welfare benefits’. Similarly, discrimination in the labor market might be responsible for a large part of Surinamese unemployment rates. But the question remains if discrimination that much more influential in Amsterdam to explain the far higher unemployment rates? From chapter 5, discrimination is part of both labor markets and it is hard to say that there is more discrimination in one city than the other. The literature does not support this assumption. Structural changes in the labor market have had a larger impact on the opportunity structure in Amsterdam: there is less job growth than in New York City, the percentage of immigrants in the city is higher, and general unemployment rates are higher. Overall, these differences between the cities do not seem as large as the differences in unemployment percentages for the two groups. More likely, labor market conditions and discrimination interact with welfare state benefits. If the labor market is not favorable, and discrimination excludes some groups more than others, these groups might ‘give up’ in the search for jobs. Economic restructuring and an increase in bifurcation in the labor market can explain how unemployment levels are highest for those with little education, and lowest for those with a bachelor’s degree or higher. Additional, unemployment levels for both groups are far higher than the general unemployment rate in each city. This statement holds for each educational level, although the higher the education, the less the difference with the unemployment rate of the city. It is a fair assumption that city unemployment levels differ per educational group for the general population too, 221 implicating that for higher educated Puerto Ricans and Surinamese, unemployment levels exceed those for non-minorities at the same education level. Even without that assumption, the group-levels remain higher, and the only plausible explanation is that discrimination keeps the groups from participation. This combination of factors is even more plausible when unemployment rates and occupational levels are combined in the investigation of changes in occupational positions of the two groups. Figure 10-3: Group, educational level, and change over time in occupational levels at highest level of education = less than high school 8.00—J group of study 7 00-1 —Surinarnese - Puerto Rican o» 8 l 5.00-1 8 1 3.00~ \ 2.00- % Estimated Marginal Means 0.00— I l 1.00 2.00 time period of study 222 Estimated Marginal Means Estimated Marginal Means at highest level of education = high school completed 8 1 7.00% 6.004 8 L 4.00— 8 1 2.00— i 0.00—l group of study ———Surinamese Puerto Rican l l 1.00 2.00 time period of study at highest level of education = some post secondary 8.00— 7.00~ 6.00- 5.00— 4.00—1 3.00- 2.00— 1.00—— 0.00— group of study —-6urinamese --~ Puerto Rican l l 1.00 2.00 time period of study 223 l Means rna Estimated Marg at highest level of education = BA and higher 8.004 7.00—~ 8.00J 5.00— 4.00- 3.00— 2.00-— 1.00— 0.00— group of study ————6urlnamese -——Puerto Rican M I l 1.00 2.00 time period of study On level 1 of education, the Puerto Ricans had a slight increase in occupational standing: Surinamese position decreased on this level. On the second educational level, the changes for both groups were very similar: there was a relative lack of changes. On level three (some post secondary but no bachelor’s degree), the Puerto Rican graph is similar to the changes on level 1, the Surinamese again experiences a decrease in occupational position. The highest education level shows where the improvement of occupational position occurred for the Surinamese. Again, the Puerto Ricans change is pretty stable, as for the other levels of education. Thus, overall, while the occupational improvements for Puerto Ricans are pretty stable over each educational category, there are large differences per category for the Surinamese. The main increase in occupational position is linked to high education levels, while the 224 |l' position for the lowest educated has deteriorated. Similarly, the position for those with ‘some college but no BA’ has gotten worse for Surinamese from 1991 to 2002. The unemployment and occupation level patterns of Surinamese in Amsterdam are in line with the notion that labor market conditions in urban areas, have shifted: while unemployment levels for Surinamese with a high school education and with ‘less than a BA’ have decreased, the shift in occupational level for those with ‘less than a BA’ is dramatic: this supports the notion of the bifurcated labor market where the majority compete for lower level jobs, and a minority with high education are concentrated in the higher end jobs. Surinamese in Amsterdam with education levels lower than a bachelor’s degree but with high school completed seem to compete for the same kind of jobs, while those with less than high school are either unemployed or working in jobs that are of the lowest categories. There is a large decrease in average occupation level for this group. For Puerto Ricans in New York City, different shifts took place. Unemployment levels slightly increased over time, and only a slight decrease for those with ‘some post secondary’ education. Occupational positions showed a similar slight increase. Overall, the picture that emerges is fairly stable. What can explain these differences between the groups in unemployment levels, changes in unemployment levels, occupation levels, and changes in occupation levels? Welfare states and Economic Mobility In the literature on economic mobility and the welfare state, it is assumed that industrialization would lead to more openness in societies. Key in the process is the 225 increased influence of education levels, as opposed to the influence of the position of the father (from chapter 4, under International Economic Mobility). A second assumption from the literature is that the political constellation of countries has an influence on the degree of openness: there are differences between rightist versus more leftist governments and differences in mobility patterns between types of welfare state. While I was not able to test these assumptions statistically, some general statements can be made. The United States and the Netherlands are industrialized societies, and have different types of welfare states. Industrialization has an effect on human capital characteristics: with industrialization, there is less effect of generational inheritance in education and occupation. Father’s (parents’) education and occupation tend to have less of an impact on education or occupation of the next generation. The differences in welfare states are expressed in the measures used in this dissertation (as discussed below) and are clear from the literature: The United States is clearly the country that comes closest to having an individualist mobility regime, where structural position and individual resources determine outcomes [. . .] The Netherlands is at the other extreme from the United States in the lack of individual-level sensitivity to structural change (DiPrete et al. 1997, p. 350-1). In the United States, there does not seem much change in outcomes. It could be argued that the researched period is too short to examine structural changes in relation to group mobility in occupational positions (or lack thereof). The changes in individual resource, in the sense of group educational levels, are similarly lacking. This could be linked to the ‘individualist welfare state’ because there are few means provided by the welfare state for change in resources. 226 In the Netherlands, educational levels have increased for the Surinamese, but occupational positions have not followed. This again can be due to the short period of time of the study. On the other hand, structural changes seem to have had an effect on occupational positions: a polarization is emerging, just as postulated by Hamnet (from chapter 5): The post-industrial societies can experience two alternative kinds of polarization. In the strong welfare states the polarization is between 'a small, but highly upgraded insider structure and a large outsider surplus population. In the other case, a large service class proletariat will constitute the pivotal source of polarization (Hamnet 1998, p. 25). The polarization can be due to both labor market changes and discrimination on the labor market. On a more positive side, the increase in education can implicate a better future for Surinamese in Amsterdam. The higher education levels might mean that in the (near) future, the ‘catching’ up with the general population can be completed. It also means that for this group, the impact of parents’ educational position is less than that for Puerto Ricans in New York City. The ‘industrialization assumption’ is confirmed in the data sets for Amsterdam, but not for those of New York City. Education levels for Puerto Ricans have remained stable. The effect of industrialization might have a combined effect with welfare state provisions in the change in educational achievements. The link between education and industrialization has been made to industrialized countries in general, but I have not seen the specific connection between differences in welfare states and the impact on (generational) shifts in education. 227 Education was one of nine explanatory variables in this research. I used education levels to control for their effect on economic mobility. I did not look at the relationship between education and the welfare state. From the literature on economic mobility, there was no indication that the effect of education can be differential for different groups. In fact, it is assumed that higher education leads to higher occupational levels (as described in chapter 4). The lack of this effect for Surinamese in Amsterdam is remarkable in that context. Migration and Mobility The conclusions on wage-parity from previous research have given different predictions for the time period in which equal wages for immigrant groups would be achieved (chapter 4). The early research that predicted 10 to 15 years for a transitional period already had faced challenges from later work, and discrimination and a changing labor market have made outcomes unpredictable. In the new global economy, the re-positioning of migrant and minority groups might take longer - if achieved at all. The caution of ‘if achieved at all’ seems to be appropriate for Puerto Ricans in New York City. Discrimination, structural changes in the labor market, and the lack of provisions (in education) have created an occupational positioning with ‘lack of change’. The results do not show any indication of a ‘more positive side’ for the position of Puerto Ricans in New York City, even though the labor market structure seems more opportunistic. While a ‘causal explanation’ cannot be derived from the statistics, the lack of change in education and occupation can almost only stem from a lack of provisions. 228 The ‘quality of the group’ (chapter 4) was the same for both groups in the 1980s: both groups had similar educational levels, similar occupational levels, and other similar characteristics (as described in chapter 7). Surinamese in Amsterdam were able to improve their educational status which indicates that ‘quality of the group’ does not mean a stable component. To say that today’s immigrant groups do not share the characteristics of earlier groups only means that they do not necessarily share those characteristics when entering the country. As the Surinamese example shows, group characteristics can (drastically) improve. Their labor market status does not seem to relate to the group characteristics. This part of immigration theory does not match. In theory, education levels are used as one of the most important aspect to estimate occupational position. From the changes in education levels of the Surinamese, one would therefore predict a larger change in occupational positions. The Dual Labor Market Theory and the Skills-Mismatch theory can shed light on the lack of change in economic positions for Surinamese. In chapter 4, the following quote was used to summarize the contemporary situation in global cities: Immigration can be seen as providing labor for: (1) low-wage service jobs, including those that service (a) the expanding, highly specialized, export-oriented service sector and (b) the hi gh-income lifestyles of the growing top level professional workforce employed in that sector; (2) the expanding downgraded manufacturing sector, including but not exclusively, declining industries in need of cheap labor for survival, as well as dynamic electronics sectors, some of which actually can be seen as part of the downgraded sector. A third source of jobs for immigrants, is the immigrant community itself (Sassen 1988, p. 22). The lack of change in occupational positions for the Surinamese could be due to these changes in the structure of the labor market in Amsterdam. Again, it also could 229 be related to the relative short period of time that I have investigated: chronologically, it makes sense that occupational positions follow changes in education levels, and that the shifts in occupations are occurring now — and will show in the data of the next SPVA. For Puerto Ricans in New York, the main increase in occupational levels was for the lowest educated: people without a high school diploma. This might simply refer to the movement out of manufacturing (categories 2 and 3) to the clerical and service sector. But it does not seem to be in line with theories on the changing labor market structure. A decrease in the level of occupational classification would be implied from theories in skills-mismatch and dual labor market. ‘Ethnic queueing’ and the shift in occupations over time would explain this shift: since Puerto Rican have a long-established tradition in New York City, this increase in levels for the least educated might come from their longer length of stay than that of other groups. Education, Unemployment. and Occupation In general, education levels for the Surinamese have increased more than those of Puerto Ricans, and unemployment levels have decreased, but remain at a higher percentage than those of Puerto Ricans in New York City. Finally, occupation levels for Surinamese have not increased as much, but remain slightly higher. The differences in these results can be linked to the structure of the labor market and to provisions of the welfare state. Surinamese have been able to increase their education levels, and most likely, the welfare state and educational provisions have had an important role. Occupation levels in the two global (or world) cities are affected by the structure of the labor market. The downsizing in manufacturing and 230 the increase of the service sector can both be seen in the occupational positioning of the groups. The decrease in occupation level for Surinamese with ‘some post secondary education’ can be linked to these changes, but labor market structures cannot account for all the changes, nor for the lack of change. The similarities in the occupational patterns and unemployment levels of the two groups can be linked, again to the changing labor market structures, but also to discrimination. Both groups have reported a high level of experiences of discrimination, and their higher unemployment levels than the general unemployment in the cities can be linked to discrimination. The labor market structures might have changed, and time might give a different picture of occupation levels. The high unemployment levels and the lack of occupational repositioning in the ‘better’ jobs seem to be attributable to discrimination on the labor market. Methodological issues Sample Size In this dissertation I encountered two problems with sample size: in the first step of the analysis, the sample sizes of the Puerto Rican data sets were ‘too large’. In the second step of the analysis, the sample sizes of the Surinamese data sets were ‘too small’. While this might seem as one problem, it was not. In the first step, one-way ANOVA could not handle the large difference in sample sizes, in the second step, the values in the original SEM model could not be estimated because of the issue of ‘over-identification’. This is a different problem than differences in sample size: the second problem is related to the number of variables needed in order to ‘add’ unknowns (such as paths between observed variables and latent factors, and paths 231 between factors). The increase in variables in turn demands an increase in sample size. The solutions to these problems were not evident. Few books or articles discuss the problem of differences in sample sizes, while this might be a problem common in international comparative research. Researchers are dependent on secondary data sets that are not (always) collected with the intention of future comparative goals. Future data collection should be done with this limitation in mind. It seems to be hopeful that there are international institutes that collect data with the goal of international comparisons. Sample size compatibility should be one of the considerations for the collection of international comparative data. Limits of the Measures From the Confirrnatory Factor Analysis, both the Labor Market Context of Reception as the Welfare State Context of Reception generated good estimates. All measures loaded high on the factors and the percentages of variance that was explained by the measures were high (tables A3 7 and A3 9, appendix 3). This means that both composite measures are good indicators of each context. Unfortunately, the third Context of Reception, the Ethnic Group Context of Reception, did not have similar good loadings or a high percentage of variance explained. The implications are two-fold. First, the measures used for the labor market and the welfare state can be used in future research. Second, the ethnic group context of reception needs redefinition of measures. I will discuss these implications below. 232 Welfare State and Labor Market Contexts of Reception The measures that were used to estimate the labor market and the welfare state do give an indication of difference and, for welfare state provisions, of what is spend. The factor loadings are high and the percentage of variance explained is high for both these ‘contexts’. In a direct comparison, where a causal assumption is presented, it might be more plausible though to look at more than two welfare states. The use of two states gives a dichotomous picture: percentages are higher or lower in one country than the other, and while the picture is interesting, it might be more informative to use more countries in a comparison. A larger range of different labor markets and/or welfare states can provide more information on their impacts. This could be used to give a picture of what impact the welfare state can have on a) labor market conditions and b) economic opportunities for various groups within each welfare state. In that way, the models and their fit on the data can be contrasted across different welfare states, and it can be linked to other factors that have an impact on economic mobility or occupational positions, as discrimination. With a two-welfare- state-model, the conclusions are (statistically) limited. Ethnic Group Context of Reception The three measurements used for the ‘ethnic group context of reception’ were not suitable to use factor analysis (and hence not suitable for Structural Equation Modeling). The measures used were all ‘categorical’ which causes problems with factor analysis. The variable ‘Female headed household’ did not load high on the factor that was created to estimate this part of the context of reception. One of the explanations for 233 the lack of fit can be that the variable was measured with only two categories (binary variable). The variable could be reclassified into percentages, to make it a continuous variable, but this step did not improve the factor loading or the variance. There is no statistical evidence that this variable can be used to create a factor that estimates the context of reception of the ethnic group. The second variable, ‘Length of stay’, was measured in categories, but could be reclassified in years, provided that the data is collected in such way: some data collections only report the length of stay in (5 year)-classes. A similar recategorization would be more problematic for the third variable used, ‘level of education’. If education would be measured in years, distortions would arise from the different education systems across countries. In an international setting, years of schooling do not overlap. One could argue that ‘diploma obtained’ is already problematic, the problems would even be larger when education is measured on years. All measures for the ‘ethnic group context of reception’ were determined by theoretical assumptions on the influence of these characteristics on mobility. Education levels, household composition, and length of stay are not aspects that define an ethnic group, but these are group-characteristics that have an impact on economic mobility of any group in any society. The notion of an ethnic group context of reception is defined by the support systems that exist within (ethnic) groups, as the prevalence of an ethnic economy. When these support systems are not available, it is difficult to create this context of reception based on variables that characterize an ethnic group. The selection of female-headed households as one of the measures is 234 based on empirical differences between groups. This is not to say that this is a stable characteristic that defines the ethnicity of the group(s), just as education level is not a stable, unchangeable characteristic. Occupational Mobility or Position A final discussion on the measures is related to ‘occupational position’. It can very well be argued that the way occupation is operationalized does not give a good portrayal of the ranking of positions in society. Issues of skill level are supposed to be combined in the ILO-classifications, but there are some points that can be made against. First, the occupational classifications are (often) related to the structure of industrial society: there are different categorizations for different types of manufacturing jobs (in the ILO case, and in this dissertation, recoded categories 2 and 3). On the other hand, the service sector does not have much differentiation. There is some stratification through the divisions into ‘elementary occupations’, clerks, and different professional jobs, but the ‘service sector’ remains a large category including a variety of occupations. The results should thus be examined with care: the increase in the percentages in the service sector (category 5) are used to indicate an improvement in occupational position, while the jobs might not be safer, cleaner, better-paying, or more complicated than employment in categories that are classified as lower. There are alternative classifications, but these do not solve for this problem. The scales presented in chapter 4 have very similar categories, and a very similar lack of divisions within the service sector. Specification and reclassification of the 235 occupational structure to fit post-industrial society, on an international scale, is needed. Discussion The main findings of this research can be summarized as ‘there is a remarkable increase in education levels for Surinamese in Amsterdam, but their occupational distribution has not followed. The position of Puerto Ricans in New York City has remained stable over time’. The main explanation for the increase in education levels can be linked to welfare state provisions, while the lack of changes in other aspects is mainly explained by discrimination on the labor market and the changing structure of the labor markets. To support these findings with statistical evidence, several steps can be taken. The first step is the improvement of the measures: the ‘ethnic group context of reception’ could not be estimated, it might be difficult to find suitable measures that can display this underlying factor. A main problem lies in different categories that countries use to measure the variables. Education could be measured in years, but there are new issues that arise with that scale. Education levels need to be part of research on occupational positioning or mobility: even though the occupational levels of Surinamese in Amsterdam did not follow the increase in education levels, most research confirms the importance of education in predicting occupation. Other measures could be included, but there will be problems with generalizability. For example, ‘knowledge of the dominant language’ could be used as a characteristic for the ethnic group and their opportunities on the labor market. But there are problems with this measure: are the consequences the same for someone who does not speak 236 Dutch (but does speak English) on the Dutch labor market as for someone who does not speak English on the US. labor market? It could easily be argued that this is not the case, a large percentage of the Dutch population speaks English. On the other hand, are the implications the same for speaking ‘Spanish only’ in New York City as the implications of speaking ‘Sranang Tongo only’ in Amsterdam? There are many different consequences linked to language-knowledge, and knowledge of the dominant language might be more important in some countries, and in some cities, than in other situations. Similarly, a measure on segregation would encounter problems of cross-national meaning. If used, segregation indices should be linked to, for example, the transportation availabilities in different countries. Another linkage should be to social distance and not just geographical distance. Overall, it might be hard to find measures that can represent the ethnic group context of reception as a factor in different countries or locations. The only solution for the inclusion of characteristics of the ethnic group might be to use these as separate variables instead of a factor on the context of reception. The welfare state and labor market contexts of reception were estimated through three measures each, that reflect the underlying factors. The inclusion of more welfare states and more labor markets will make the comparison more interesting. With more welfare states to compare, a gradation of effects can be estimated, instead of a ‘polarization’ as is done with the comparison of the United States and the Netherlands. The selection of groups and countries should be done with great caution: what are similar groups in similar labor markets in different welfare states? Can capital cities as Paris and London be included in the comparison, and are groups in 237 those cities from a similar background? While Amsterdam is officially the capital of the Netherlands, the government resides in The Hague which creates a different labor market structure. Amsterdam and New York City might not have labor market structures comparable to London or Paris. And while A1 gerians are colonial/post- colonial migrants in France, there are differences with the studied groups, most notably on religion. A final improvement in the measures should be related to changes in the post- industrial labor market. The service sector should not be treated as one entity, and divisions, as were made in the manufacturing sector, are needed. Skill levels differ, job security is variable, and the social standing of different service jobs needs to be included. In a post-industrial society ‘soft skills’, that are generally not measured in human capital characteristics, might become more important. The ability of employees to interact, both with clients as with others, is more important (or is perceived as more important) in a service economy. The measurement of soft skills, and the measurement of the perception of employers of soft skills of possible employees, should be included in the opportunity structure of the post-industrial society. If possible, some measure of income should be attached to the categorization of occupations. This might be difficult to achieve, and should account for standards of living, for other forms of income, and for tax-levels, deductions, and other ways states redistribute the monetary flows. Finally, ‘wealth’ and ‘income from wealth’ are not included in national or international measures of economic mobility. As with income and other forms of redistribution in society, it might be difficult to include a cross-societal measurement of wealth and the accumulation of wealth it generates. 238 I was not able to look directly at the influence of the welfare state on occupational positions but I could derive some conclusions on the possible impact, related to other aspects of the contexts, as changes in the labor market structure and discrimination. Another interesting relation, the increase of education levels of Surinamese in Amsterdam could use further elaboration. How can this increase be explained, and are there alternatives to the explanation that I have thought most plausible, the influence of the welfare state on education? Besides these questions on the causes for the increase in educational achievement, other important questions relate to the lack of occupational changes. In general, education is seen as one of the main causes of occupational mobility in industrialized society. Is this the same for post-industrial society, and if so, why has that effect not translated into better occupational positioning for Surinamese in Amsterdam? The complex statistical model could not capture the complex reality of two groups in two labor markets in two welfare states. Instead of a confirmatory dissertation with causal relations on economic mobility of migrant groups in two welfare states, this research has become an exploratory analysis of problems and issues in international comparisons of group mobility. The cautions and alternatives can generate additional models and possibly a causal assessment of the influence of the welfare state, the labor market structure, and the characteristics of the ethnic group on economic positioning and mobility across societies. 239 Appendix 1: Surinamese Education System http://taalunieversum.org/onderwijs/termen/z on this website an overview of the Surinamese education system is given, It differs slightly from the Dutch system but the base is the same. Education starts with pre-school (not mandatory) and then with ‘regular lower education’ (glo). For those with extra needs, there is BSD and VSO: special lower education and continued lower education. After elementary school, at age 12, multiple educational options are available: MULO prepares students for the track towards university (through VWO to universiteir). LBGO, LTO, and LNO (lower vocational trainings) are all tracks that prepare students in vocational training, while ETO, BBQ, and ETS are simpler forms of this education. (EBO: Simple Vocational Training). After any form of Lower Vocation Training can lead to continuous education to HBO (college), Teachers’ education (IOL, Pedagogische Academies), and technical (NATIN) or economic/administrative education (IMEAO). 240 Appendix 2: Data Computations Amsterdam Job growth in Amsterdam Table A2 1: Employment growth Amsterdam 1987-1988 1984 1985 1986 1987 1988 # jobs 310000 323000 330000 334000 340000 Job growth 13000 7000 4000 6000 Growth % 4 2 1 2 Table A2 2: Employment growth Amsterdam 2001-2002 % Job Population Population Job growth Growth as % of on total total population Amsterdam-Centrum -1 78946 1 1 -.1 1 Westpoort 2.5 373 .05 .00125 Westerpark 2.4 33956 5 .12 Oud-West 4.5 31883 4 .18 Zeeburg 9.7 38692 5 .485 Bos en Lommer -1 30660 4 -.04 De Baarsjes -l4.5 34978 5 -.725 Amsterdam-Noord 5 869 10 1 2 60 Geuzenveld/Sloterm. -4.5 39842 5 -.225 Osdorp 5.7 43639 6 .342 SIoterv-/Overt. Veld 3 44034 6 .18 Zuidoost -.5 8481 1 12 -.06 Oost/Watergraafsmeer 1 57666 8 .08 Amsterdam Oud-Zuid -1 82934 11 -.11 Zuideramstel 6 46004 6 .36 total 735328 100 1.08% Ethnic Group Context of Reception Recodes Age ofjopulation Following the definitions of ILO (International Labour Organization), the population of working age is defined as including age 15 but excluding age 65 (but, as explained in chapter 8, the age is adjusted to the minimum age of 19 to make comparisons more equal: only in the Puerto Rican data sets were there respondents younger than 19, and a majority of these respondents reported ‘unemployment/not in the labor market’ while they were still in school. To avoid an overestimate of not- active people in the Puerto Rican groups, all below age 19 are excluded. 241 Migration year (StayCensg) In the 1990 census (YRIMMIG) the question on migration to the United States was described as coded into intervals. The output from the data did not show these intervals (nor did the labeling of the variable in the data set describe the intervals). Table A2 3: Year ofimmigration: Valid N/A 1949 (before 1950,1980 and 1990); 1945-1949,1970) 1959 (1950-1959,1980 and 1990; 1955-1959,1970) 1964 (1960-64,1970 and 1980 and 1990) 1970 (1965-1969,1980 and 1990; 1965-1970,1970) 1 974 (1970-74,1980 and 1990) 1980 (1975-1979,1990; 1975-1980,1980) 1981 (1980—81,1990) 1984 (1982-84,1990) 1 986 (1985—86,1990) 1990 (1987-90,1990) Total Because these numbers do not give much information, I have relied on the variable ‘years in the US’: Table A2 4: Years in the United States, intervalled (1990) 1 0-5 years 2 6-10 3 11-15 4 16-20 5 21+ 6 born in (was coded O) The value (0) was labeled N/A, these respondents are born on the mainland. 1 have recoded (0) to ‘born’, with a value of 6 (a similar category has been applied on the SPVAs). For the SPVA 1991, ‘verblkl’ (verblij fsduur in klassen) is a similar measure, but the intervals are different. These are recoded into those of the Census (since the SPVA had smaller intervals, these are easier combined). 242 Table A2 5: Categories of length of stay in the 1991 SPVA < 1 year = 1-2 year = 2-5 year = 5-10 = 10-15 =3 15-20 =4 20-25 =5 more than 25 = “\IQL’IAWNl—A The SPVA 2002 used different intervals: Table A2 6: Categories of length of stay in the 2002 SPVA Min. 25 years 20-24 years 15-19 years 10-14 years 5-9 2-4 Max. 2 years \IONUIAUJNr—i In order to avoid problems of re-categorization, I have used the non-classified variable of ‘verblijf to measure length of stay. This variable simply uses the number of years, and I have reclassified those in above intervals (this option is not available in the 1991 SPVA). For both years of the SPVA, I combined the ‘length of stay categories’ with country of birth: in the census N/A is applied to those who did not make the move form the island to the mainland, a similar category is needed for Surinamese in Amsterdam. For the Surinamese in 2002, some missing values have been added manually by cross-checking with migration year. Education (MaxEduc) The categories that were given in both the 1990-census as the 2000-census have been collapsed into categories (see chapter 7 ). The two censuses have used different questions, but the same categories. In the Census, the categories are very detailed. The first 9 categories are all including some grade before the high school diploma. Category 10 is high school graduate with a diploma obtained. Categories 11 though 13 apply to some college but degrees below a bachelor’s, and the final categories are bachelor’s degree and higher. The recodes are as followed: 243 Table A2 7: Education levels Census: Category New category 1-9 1 10 2 1 1-13 3 14-17 4 Table A2 8: Education levels SPVA 1991: from Maxdipl (highest diploma obtained) SPVA value Education level Recoded value 0 None 1 1 L0 1 2 LBO 2 3 MAVO 2 4 MBO 3 5 HAVO 3 6 VWO 3 7 HBO 4 8 WO 4 Table A2 9: Education levels SPVA 2002: from Maxdipl SPVA value Education level Recoded value None 1 BAO VBO MAVO MBO HAVO/VWO HBO WO \IONUIbWNI-‘O ##WWNNH Female headed households (FemHead) SPVA 2002: is created by combining gender (2) and family composition (values 1 and 7 are single headed households). Female-headed households were labeled 1, all other 0. SPVA 1991: combined 1 and 7 from family composition: female-headed households were labeled 1, all other 0. Census 1990: poploc: value of 0 means no father: recoded into female-headed household by making all where no father present into 1, all others 0. This is in combination with marital status, where, with the exclusion of ‘married, spouse present’ (1 ), all other values indicate that the spouse (father) is not present. This is in order to avoid labeling a household ‘female-headed’ when there is a new husband in the household. A final addition was that if a male was present in the household (boyfriend, new spouse etc), this household was not labeled as female headed. The 244 Dutch data do not make a difference between cohabitating couples and married couples. Census 2000: poploc combined with marital status gives a slightly lower percentage of female-headed households for Puerto Ricans in New York City, combined with the presence of a male partner in the household. The Welfare State: Public Health Expenditures Table A2 10: Total and public expenditure of health (US dollars calculated using PPPs) Public expenditure on Total expenditure on health health per capita per capita 1990 2000 2003 1990 2000 2003 Netherlands 965 1425 1856 1438 2259 2976 United States 1085 2008 2503 2738 4539 5635 From: hnpzl/ocdep4.siteintemet.com/pub]ications/doifiles/30200601 1P1-09-O3-01- t01.xls Table A2 11: Percentage public health expenditures as a percentage of total health expenditures 1990 2000 2003 Netherlands 67 63 62 United States 40 44 44 The Dependent Variable: Occupation Categories DiPrete (1997) relied on occupational categories defined by ILO in the lntemational Standard Classification of Occupation in 1968. These classifications have been modified into the ISCO-88 (1988), combining the ISCO-classification with ‘level’ of the profession. A second step in the recoding is the exclusion of categories (0) and (6): there are very few (to none) respondents that report being employed in these sectors (in 2000, only 7 Puerto Ricans answered to be employed in the military, all the other data sets do not include any respondents in this sector. Category (6) had less than .3 percent of respondents in the Puerto Rican data set in 1990, all other sets contained even less respondents). The exclusion of category 6 also implies that categories 7, 8, and 9 are recoded: in order to keep the ordinal scale for statistical purposes. I have also ‘flipped’ the codes to make a lower number the lower occupational category, and higher numbers represent the more prestigious, higher occupations. 245 Table A2 12: Occupational Categories: ISCO 88 (1) (2) (3) (4) (5) (6) (7) (8) (9) (0) Census 1990 003 through 022 023-037; 043-199 203-259 303-391 263-274; 283-290; 413—447; 456-469 473-499 503 -699; 866-874 703-865 275-278; 357; 403-408 433-435; 443; 448-455 875- 890 (military: 980- 983) 992 unemployed Census 2000 I through 33 34-186; 200-326 190-196; 286; 290,296,330- 373; 375-395; 900-904 500—593 374; 400410; 430471; 475- 494 600-613 620-762; 780- 785; 874-890 770-775; 790- 873; 912-960 41 1-416; 420- 425;472;495- 496; 961-975 (military: 980- 983) 992 unemployed SBC 1992* 93, 98 55 56 62 63 69 73 75 76 77 78 79 82 83 84 85 88 89 9192 95 96 97 492 493 53 58 59 64 65 66 67 68 7172 74 86 87 314 315 484 513 514 515 2123 25 29316 317 34 37 4142 43 471 473 485 494 495 516 517 518 54 57 24 44 262 263 268 271 460 461 462 463 466 467 468 472 521 264 265 266 267 272 28 464 465 482 483 11261 Occu ISCO Occu Final From: http://www.ilo.orgjpublic/english/bureau/stat/isco/isco88/publ3.htm: by skill level (the higher the number, the more complicated skills are required), the major groups are in parentheses). *The SPVA uses codes by the CBS (Centraal Bureau voor de Statistiek, Central Bureau of Statistics). The SPVA-91 used occupational codes from 1984, these have been recoded to match the Standaard Beroepenclassificalie 1992 (SBC92: Standard Occupational Classification). The categories here are simplifications, I can provide more detailed overviews. All numbers have 5 digits, the numbers displayed are the first two or three of the total. There is a general label in the SPVA92, coded ‘20000’. I have added this label to group 5: there are specific numerations for occupational categories 26*", 27*”, 28*", and 29*", indicating that the others are part of the lower 2000O-numbers. Since the only other ISC088 category containing a 2“" 246 number is the farming category, it is safe to assume the occupations match the ISC088 category 5, service industry. The one person in the SPVA91 that was assigned an occupational code of ‘46000’ has been relabeled to category 7. Transformation of the Dutch occupation categories: The recoding of the SPVA91 categories from occupational categories of 1984 to the SBC92 was done with the help of lace Dagevos of the SCP (Social and Cultural Flaming Bureau) who was kind enough to sent me the SPSS syntax file that was created by him to make the transition. I then made a file containing the responses of both years (linked to the year of the SPVA) and deleted the last two digits of each code (to simplify the transformation process). Then, using SPSS, I recoded the labels of the dataset into the 9 categories of the ISCO-88. I created one file of occupations for both 1991 and 2002, and excluded those who did not report to be employed (see below under ‘unemployment’). Transformation of the Census occupation categories: This was done through SPSS, excluding 992: the degrees of unemployment are discussed separately (all other values are labeled as system missing, except for military occupations: these were recoded into 0 — in 1990, none reported to be occupied in military occupations, in 2000, only 7 people reported to have a military job. These 7 people are excluded from the analysis). Since the two census files contain different occupational codes, the transformations had to be done first, then one file could be constructed with the new categories (labeled OchSC088 and ochSCOOZ) A category ‘unemployment’ The unemployment rates are computed by recoding all occupational categories (except for ‘military’ in 2002: these cases are deleted) into 1, and others as zero. To compute the frequencies in ‘occupation’, I had to make an adjustment for the SPVA91: in 1991, the survey question referred to ‘last job/current job’ thus including people that reported their last job, but were currently (in 1991) not employed. The variable ‘nu werk ’, or ‘employed now’, that had the answer categories ‘yes’ (1) and ‘no (2) and was combined with ‘last job/current job’ to exclude those who did not have a ‘current job’. In 2002 the question only referred to ‘current job’ and the adjustment was not needed. Summary The last table of this appendix summarizes the measured used. All measures are displayed in the table: 247 Table A2 13: Measures used in the dissertation: Variable Label Values Year of study year 1990, 1991, 2000, 2002 Group of study group 1=Surinamese 2=Puerto Ricans Length of stay StayCensus l=0-5 years 2=6-10 3=11-15 4:16-20 5=21+ 6=bom in Maximum diploma obtained MaxEduc l=less than high school 2= high school diploma 3= some postsecondary 4=BA and higher Head of household F emHead 0=other l=female headed Percentage immigrants in the perclrnmi Percentage crty Percentage job growth in the JobGrowth Percentage City Unemployment level in the CityUnempl Percentage crty Social expenditure as % of SocExpen Percentage GDP Taxes on work as % of total TaxWork Percentage labor cost Public expenditures as a PublicTotal Percentage percentage of total (on health) Unemployment levels of the UnEmploy 0=employed; 1= roups unemployed Occupation levels of the Occu—final Categories: see table 8-1 group 248 Appendix 3: Steps in the analysis Step One: ANOVA For the first step, answering the question ‘are there significant changes in the economic position of each group?’, the positions in time were contrasted. To see if one group has a significantly larger change in position, tests of the significance of the differences between the groups were done. Results of the one way-ANOVA The results of the one-way ANOVA were seriously influenced by sample size: because the Puerto Rican data-set are about 1000 times as large as the Surinamese data-set, a slight change in their position, on education, on occupation, is significant. The most illustrative results from the one-way are the 95% confidence intervals: even though the change in the mean for Puerto Ricans is significant over time, they show that the change is relatively small, as compared to the changes in mean-values for Surinamese. I report some of the findings of the one-way ANOVA, to illustrate the problem of the difference in sample size with regards to education. The second illustration is on employment, where I display the 95% confidence intervals. Education Tables A3 1 and A3 2 Give the results of one-way ANOVA tests in the changes in education levels for Surinamese in Amsterdam and for Puerto Ricans in New York City. The tables give the F-value for the significance of the change per group: a higher F-value means that the change is more significant. These tables illustrate the influence of sample size: in chapter 9, the changes in education are displayed graphically, and it appears to be obvious that the improvements are greater for the Surinamese. The one-way ANOVA does not confirm this result, due to the differences in sample size. Table A3 1: One-way ANOVA: changes in educational levels for Surinamese in Amsterdam 1991-2002: Highest level of education Sum of Sgarss df Mean Square F Sig. ‘ Between Groups 26.609 1 26.609 27.866 .000 Within Groups 344.713 361 .955 Total 371 .322 362 249 Table A3 2: One-way ANOVA: changes in educational levels Puerto Ricans 1990-2000 Highest level of education Sum of Squares df Mean Square F 8L Between Groups 363.681 1 363.681 386.488 .000 Within Grouros 36304.352 38581 .941 Total 36668.034 38582 Employment As an illustration, table A3 3 shows the 95% confidence intervals for ‘unemployment’. The higher the mean, the more unemployment: since there are only two values, 0 and 1, this is the only way to interpret the table. The ranges of Puerto Ricans are very close together, which makes sense when we look at the percentages given in the chapter. The ranges of the Surinamese are far larger, again, sample size impacts the clarity of the data. While the change in Puerto Rican unemployment is a decrease of 2 percent on 31 percent, the change for Surinamese unemployment is a decrease of 13 percent on 47 percent. Table A3 3: Confidence Interval for ‘unemployment’, both groups over time. 95% Confidence Interval for Mean Mean Lower Bound Upper Bound Surinamese 1991 .4694 .3989 .5399 Surinamese 2002 3352 .2654 .4050 Puerto Ricans 1990 .3130 .3065 .3194 Puerto Ricans 2000 .2931 .2866 .2996 In order to avoid the strong deterministic impact of sample size I have run two- way ANOVAs: this method of analysis uses all groups to compare and contrast and is not as sensitive to differences in sample size. For a two-way ANOVA, all the data has to be in one set. Occupation Table A3 4 displays the count and percentages of Puerto Ricans and Surinamese in. each occupational category per year of study. 250 Table A3 4: Distribution of Surinamese and Puerto Ricans per occupational category, over time. 0mg _tjr_1al Total I 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 72000 1958 1201 1144 1844 3061 1426 2280 406 13320 % 14.7% 9.0% 8.6% 13.8% 23.0% 10.7% 17.1% 3.0% 100.0% 2002 14 2 4 39 27 11 19 3 119 % 11.8% 1.7% 3.4% 32.8% 22.7% 9.2% 16.0% 2.5% 100.0% 1991 12 3 9 37 19 4 18 2 104 % 11.5% 2.9% 8.7% 35.6% 18.3% 3.8% 17.3% 1.9% 100.0% 1990 2030 1837 135 2261 3238 805 1364 675 13562 % 15.0% 13.5% 10.0% 16.7% 23.9% 5.9% 10.1% 5.0% 100.0% Results from the two-way ANOVA Educaflon The changes in educational levels are discussed in chapter 9 and 10. The additional graphs elaborate on the results: these graphs show the differences in the slopes: the changes in educational levels have different slopes per group. If the group slopes differ, that means there is ‘interaction’: there is an interplay between ‘group’ and ‘year’ in this case. From the test, it was revealed that the interaction is significant: that means that the changes in educational levels were different per group over time. 251 Figure A3 1: Educational levels per group, over time Estimated Marginal Means 4.00— 1300—4 2.00- 1.00-1 Estimated Means: highest level of education ./ f 1990 Employment Employment is measured through 1 and 0 coding, and the higher the number, the higher the unemployment levels (unemployed is coded as 1). The number can be 1 l 2000 1991 year of study I 2002 group of study —Surinamese ~+Puerto Rican interpreted as percentages. Table A3 5 shows the significance of the interaction between ‘group’ and ‘time’ and unemployment levels, confirming that the levels changed differently for the two groups: Table A3 5: Tests of Between-Subjects Effects: Dependent Variable: number of unemployed Type 111 Sum Mean Source of Squares df Square F Sig.__ Corrected Model 9.357(a) 3 3.119 14.749 .000 Intercept 184.386 1 184.386 871.949 .000 year‘sroup 9.357 3 3.119 14.749 .000 Error 8237.359 38954 .211 Total 1 1853.000 38958 Corrected Total 8246.716 38957 a R Squared = .001 (Adjusted R Squared = .001) The interaction between year" group again is significant, and looking at the plot it makes sense: 252 Figure A3 2: Graph of the ‘mean level’ of unemployment per group per year Estimated Means of number of unemployed 1.00—J| group of study a, ' ————6urinamese 5 -~ Puerto Rican g 0.75— 76 .E g 050—. 2 l \ 'o 2 0 25 re , —- .§ H in 1.1.1 0.00—4 ‘7‘ I I "I‘ I 1990 2000 1991 2002 year of study Step Two: Context Initially, the influences of the different contexts were going to be assessed by using Structural Equation Modeling. Due to several issues, this is not possible (chapter 9). First, the model needed to allow for covariance between the government or welfare state context of reception and the labor market changes and opportunities. With the inclusion of more paths, there were identification problems in the model. One way to solve for identification problems is through the increase in (observed) variables, or, to increase the measures per context. This in turn would create problems with sample size, the increase in variables demands an increase in sample size for the Surinamese sample. An alternative approach is to use factor analysis, and generate regression coefficients for the factors. I run several analyses, and, as described in chapter 9, these did not generate a fitting solution. The first, basic run using factor analysis in SPSS revealed that the variance explained in the ‘ethnic group context of reception’ was low: table A3 6 gives the results: 253 Table A3 6: Total Variance Explained Ethnic Group Factor Initial ggenvalues Extraction Sums of Squared Loadipgs_ % of Cumulative Total % of Variance Cumulative % Total Variance % 1 1.112 37.070 37.070 1.112 37.070 37.070 2 .994 33.142 70.212 3 .894 29.788 100.000 Extraction Method: Principal Component Analysis. For the two other factors, the amount of variance explained is acceptable/good (Tables A3 7 and A3 8): Table A3 7: Total Variance Explained Labor Market Initial Eigenvalues Extraction Sums of S uared Loadings % of Total % of Variance Cumulative % Total Variance Cumulative % 1 2.766 92.214 92.214 2.766 92.214 92.214 2 .190 6.327 98.541 3 .044 1.459 100.000 Extraction Method: Principal Component Analysis. Table A3 8: Total Variance Explained Welfare State lnilialgEigenvalues Extraction Sums of Squared Loadings % of Total % of Variance Cumulative % Total Variance Cumulative % 1 2.597 86.583 86.583 2.597 86.583 86.583 2 .393 13.086 99.669 3 .010 .331 100.000 Extraction Method: Principal Component Analysis. Of the three variables that were assumed to measure the ethnic group context of reception, ‘female-headed households’ does not have a high loading on the communality (or, on the common aspect of the factor). The coefficient for ‘female- headed households’91 is relatively low: 9‘ 1 have recoded the variable ‘female-headed household’, which was a Dummy-variable with only the values (1) and (0) into a percentage value, using the results of the first step of analysis. This did not improve the model, the original variable is used in the final steps. 254 Table A3 9: Factor loadings for Ethnic Group Context of Reception: Loadinfln Factor highest level of education female-headed household Years in country .731 -.241 .721 Extraction Method: Principal Component Analysis. The alternative, more appropriate method of extraction, Unweighted Least Squares, did not improve the results: Table A3 10: Total Variance Explained for the factor Ethnic group context of reception, using ULS Initial Eigenfllgs Extraction Sums of Squared Loadin‘ggd °/o of Cumulative % of Cumulative Total Variance % Total Variance % 1 1.112 37.070 37.070 .224 7.481 7.481 2 .994 33.142 70.212 3 .894 1 29.788 100.000 Table A3 11: Loadings using ULS Factor kw- a _ Years in country .280 highest level of education 373 female—headed household -.057 The data and variables do not match the theoretical assumptions of three contexts of reception. The results from the factor analysis do not legitimate the use of factors. I could force the variables in three factors, which gave a regression model with an R- square of .161, a severely low estimate (and the coefficients were significant again: this goes back to the large sample sizes). Table A3 12: Model Summary: estimate of R Square J Adjusted R Std. Error of Model R R Square Square the Estimate | 1 | .401(a) .181 .161 1.89120 | 255 Table A3 13: Coefficients(a) of the regression Unstandardized Model Coefficients Standardized Coefficients 8 Std. Error Beta t SiL 1 (Constant) 4.033 .012 343.379 .000 REGR factor score Ethnic 1.827 .026 .396 70.536 .000 group REGR factor ' score Labor .050 .013 .024 3.960 .000 Market REGR factor score Welfare .031 .014 .013 2.203 .028 State The only other way to measure the influence of the different contexts of reception is through a regression analysis, but as explained in the first step, this would counter severe problems of multicollinearity. 256 References Agresti, Alan, and Barbara Finlay. 1997. Statistical Methods for the Social Sciences. Upper Saddle River, NJ: Prentice Hall. Alba, Richard D., John R. Logan, and Brian J. Stults. 2003. "Residential Inequality and Segregation in an Immigration Era: An Analysis of Major US. Metropolitan Regions in 1990." Pp. 119-148 in Host Societies and the Reception of Immigrants, edited by Jeffrey G. Reitz. La J 011a, CA: Center for Comparative Immigration Studies University of California, San Diego. Alba, Richard, and Victor Nee. 2003. Remaking the American Mainstream. Assimilation and Contemporary Migration. Cambridge: Harvard University Press. Alderson, Arthur S., and Francois Nielsen. 2003. "Globalization and the Great U- Turn: Income Inequality Trends in 16 OECD Countries." in Globalization and Society. Processes of Diflerentiation Examined, edited by Raymond Breton and Jeffrey G. Reitz. Westport, CT: Praeger. Amersfoort, H. van. 1968. Surinamers in de [age landen - De ontwikkelt'ng en problematiek van de Surinaamse migratie naar Nederland. The Hague: Staatsuitgeverij. —-——. 1992. "Ethnic residential patterns in a welfare state. Lessons from Amsterdam, 1970-1990." New Community 18:439-456. Amersfoort, Hans van. 1998. "Governments and Immigrants: Similarities and Dissimilarities in the Western European Experience. ." Pp. 119-136 in Regulation of Immigration. International Experiences. , edited by Anita Becker, Kees Groenendi j k, Tetty Havinga, and Paul Minderhoud. Amsterdam: Het Spinhuis Publishers. Amersfoort, Hans van, and Cees Cortie. 1996. "Social polarization in a welfare state? Immigrants in the Amsterdam region." New Community 22:671-688. Amersfoort, Hans van, and Mies van Niekerk. 2006. "Immigration as a Colonial Inheritance: Post-Colonial Immigrants in the Netherlands, 1945-2002." Journal of Ethnic and Migration Studies 32:323-346. Amersfoort, Hans van, and Rinus Penninx. 1998. "Western Europe as an immigration area." Pp. 42-68 in International Migration. Processes and interventions. , edited by Hans van Amersfoort and J eroen Doomernik. Amsterdam: Het Spinhuis. 257 Anderson, Benedict. I991. Imagined Communities. London/New York: Verso. Anderson, James C., and David W. Gerbing. 1987. "Structural Equation Modeling in Practice: A Review and Recommended Two-Step Approach. " Quantative Methods in Psychology (online). Barrington-Moore, Jr. 1966. Social Origins of Dictatorship and Democracy. Lord and Peasant in the Making of the Modern World. Boston: Beacon Press. . Barry Figueroa, Janis. 1991. "A Comparison of Labor Supply Behavior among Single and Married Puerto Rican Mothers." Pp. 183-202 in Hispanics in the Labor Force. Issues and Policies, edited by Edwin Mele’ndez, Clara E. Rodriguez, and Janis Barry Figueroa. New York: Plenum. Bean, Frank D., and Gillian Stevens. 2003. America 's Newcomers and the Dynamics of Diversity. New York: Russell Sage Foundation. Beeghley, Leonard. 2005. The Structure of Social Stratification in the United States. Boston: Pearson Education. Beer, J. de. 1998. "Migratie naar Nederland." in Etnische minderheden en de multiculturele samenleving, edited by Rinus Penninx, Henk Munstermann, and Han Entzinger. Groningen: Wolters-Noordhoff. Bendix, Reinhard. 1964. Nation-Building and C itizenship. Studies of our Changing Social Order. Berkeley: University of California Press. Binder, Frederick M., and David M. Reimers. 1995. All the Nations under Heaven. An Ethnic and Racial History of New York City. New York: Columbia University Press. ,Blalock, Hubert M. 1960. Social Statistics. New York: McGraw-Hill Book Company. Blau, Peter M., and Otis Dudley Duncan. 1967. The American Occupational Structure. New York: John Wiley and Sons, Inc. Booker, Anita, Kees Groenendijk, Tetty Havinga, and Paul Minderhoud (Eds). l. 998. Regulation of Migration. International Experiences. Amsterdam: Het Spinhuis. Bohning, W.R. 1998. "Top-end and bottom-end labour import in the United States and Europe. Historical evolution and sustainability." Pp. 72-85 in International Migration. Processes and interventions, edited by Hans van Amersfoort and J eroen Doomernik. Amsterdam: Het Spinhuis Publishers. 258 Boissevain, Jeremy, and Hanneke Grotenbreg. 1988. "Culture, structure, and ethnic enterprise: the Surinamese of Amsterdam." Pp. 221-249 in Lost Illusions. Carribean minorities in Britain and the Netherlands, edited by Malcolm Cross and Han Entzinger. London: Routledge. Bolle, Willeke, Henk van Dijk, and Dieke Hetebij. 1978/ 1979. "Discriminatie bij het verhuren van kamers aan gastarbeiders." Pp. 78-90 in Omdat zij anders zijn. Patronen van rasdiscriminatie in Nederland, edited by Frank Bovenkerk. Meppel/Amsterdam: Boom. Borjas, George J. 1987. "Self-Selection and the Earnings of Immigrants." The American Economic Review 77:531-553. —. 1988. International Differences in the Labor Market Performance of Immigrants. —-—. 1990. Friends or Strangers. The Impact of Immigrants on the US. Economy. New York: Basic Books. —. 1991. "The Economic Impact of Immigration. Immigrants in the US. Labor Market: 1940-1980." AEA Papers and Proceedings 81:287-291. —. 1992. "Ethnic Capital and Intergenerational Mobility." Quarterly Journal of Economics 107:123-150. —. 1994. "The Economics of Immigration." Journal of Economic Literature 32:1667- 1717. —. 1999a. Heaven 's Door. Immigration Policy and the American Economy. Princeton, NJ: Princeton University Press. —. 1999b. "Immigration and Welfare Magnets." Journal of Labor Economics 17:607-637. Bovenkerk, F. 1975. Emigratie uit Suriname. Amsterdam: Antropologisch- Sociologisch Centrum, Universiteit van Amsterdam. Bovenkerk, Frank. 1978/ 1979a. "Rasdiscriminatie bij misdaadverslaggeving in de pers." Pp. 119-1 35 in Omdat zij anders zijn. Patronen van rasdiscriminatie in Nederland, edited by Frank Bovenkerk. Meppel/Amsterdam: Boom. —. 1978/ 1979b. "Rasdiscriminatie in Nederland?" Pp. 9-30 in Omdat zij anders zijn. Patronen van rasdiscriminatie in Nederland, edited by Frank Bovenkerk. Meppel/Amsterdam: Boom. 259 Bovenkerk, Frank, and Elsebeth Breuning-van Leeuwen. 1978/1979. "Rasdiscriminatie en rasvooroordeel op de Amsterdamse arbeidsmarkt." Pp. 3 l -57 in Omdat :ij anders zijn. Patronen van rasdiscriminatie in Nederland, edited by Frank Bovenkerk. Meppel/Amsterdam: Boom. Braverman, Harry. 1974. Labor and Monopoly Capital. The Degradation of Work in the Twentieth Century. New York: Monthly Review Press. Brochmann, Grete. 1998. "Controlling immigration in Europe. Nation-state dilemmas in an international context." Pp. 22-41 in International Migration. Processes and interventions, edited by Hans van Amersfoort and J eroen Doomernik. Amsterdam: Het Spinhuis Publishers. —. 1999. "The Mechanisms of Control." Pp. 1-27 in Mechanisms of Immigration Control: A Comparative Analysis of European Regulation Policies, edited by Grete Brochmann and Tomas Hammar. Oxford/New York: Berg. Brown, Timothy A. 2006. Confirmatory Factor Analysis for Applied Research. New York/London: The Guilford Press. Burgers, Jack. 1999. "Illegale vreemdelingen en migratiebeleid." Pp. 262-284 in De ongekende stad I. Illegale vreemdelingen in Rotterdam, edited by Jack Burgers and Godfried Engbersen. Amsterdam: Boom. Butcher, Kristin F. , and David Card 1991. "Immigration and Wages: Evidence from the 1980's." American Economic Review 81 :292-296. Calavita, Kitty. 1994. "US. Immigration and Policy Responses: The Limits of Legislation." Pp. 55-82 in Controlling Immigration: A Global Perspective, edited by Wayne A. Cornelius, Philip L. Martin, and James F. Hollifield. Stanford CA: Stanford University Press. Castles, Stephan, and Mark J. Miller. 1993/1998. The Age of Migration. International Population Movements in the Modern World. : MacMillan Press Ltd. Chiswick, Barry R. 1978. "The Effects of Americanization on the Earnings of Foreign-born Men." Journal of Political Economy 86:897-921. —. 1979. "The Economic Progress of Immigrants: Some Apparently Universal Patterns." Pp. 357-399 in Contemporary Economic Problems, edited by William Fellner. Washington DC: American Enterprise Institute for Public Policy Research. Choenni, August 0. 1997. Veelsoortig Assortiment. A llochtoon ondernemerschap in Amsterdam als incorporatietraject 1965-I995. Amsterdam: Het Spinhuis. 260 Choenni, Chandersen ES, and Artwell C. Cain. 1995. "The Intensification 0f Racism in the Dutch Labour Market." Pp. 511-535 in Racism and the Labour Market: Historical Studies, edited by Marcel van der Linden, Jan Lucassen, Dik van Arkel, Els Deslé, Fred Goedbloed, and Robert Kloosterman. Bern: Peter Lang AG, European Academic Publishers. Colijn, Ineke, and Hans Smit. 1994. "The labour market and minorities in the nineties." Pp. 135-147 in Immigrant Ethnic Minorities in the Dutch Labour Market. Analyses and Policies, edited by Han Entzinger, Jacques Siegers, and Frits Tazelaar. Amsterdam: Thesis Publishers. Comelis, B. 1990. "Migratie naar Nederland." Pp. 13-30 in Etnische minderheden in Nederland, edited by HE. Entzinger and P.J.J. Stij nen. Meppel/Amsterdam/Heerlen: Boom/Open Universiteit. Cottaar, Annemarie. 2003. Zusters uit Suriname. Naoorlogse belevenissen in de Nederlandse verpleging. Amsterdam: J .M. Meulenhoff. Crok, S., K. Dignum, T. Fedorova, and J. Slot. 2002. "Amsterdamse getto's, terugkerende kwestie. Verdieping Diversiteitsmonitor 2002." edited by Bestuursdienst/MEC: O+S, het Amsterdamse Bureau voor Onderzoek en Statistiek. Cross, Malcolm, and Han Entzinger. 1988. "Caribbean minorities in Britain and the Netherlands: comparative questions." Pp. 1-33 in Lost Illusions. Caribbean minorities in Britain and the Netherlands, edited by Malcolm Cross and Han Entzinger. London: Routledge. Cross, Malcolm, and Mark Johnson. 1988. "Mobility denied: Afro-Caribbean labour and the British economy. " Pp. 73-105 in Lost Illusions. Caribbean minorities in Britain and the Netherlands, edited by Malcolm Cross and Han Entzinger. London: Routledge. Cruz, José E. 2005. "The Changing Socioeconomic Position and Political Fortunes of Puerto Ricans in New York City 1960-1990." Pp. 37-82 in Boricuas in Gotham: Puerto Ricans in the Making of Modern New York City. Essays in the Memory of Antonia Pantoja, edited by Gabriel Haslip-Viera, Angelo F alcon, and Felix Matos Rodriguez. Princeton: Markus Wiener Publishers. Dagevos, J. 2001. Perspectief op integratie. Over de sociaal-culturele en structurele integratie van etnische minderheden in Nederland. Den Haag: WRR. Dagevos, Jaco. 1998. Begrensde mobiliteit. Over allochtone werkenden in Nederland. Assen: Van Gorcum. 261 DeFreitas, Gregory. 1985. "Ethnic Differentials in Unemployment among Hispanic Americans." Pp. 127-157 in Hispanics in the US. Economy, edited by George J. Borjas and Marta Tienda. Orlando: Academic Press, Inc. Dessens, Jos A.G., Wim Jansen, Harry B.G. Ganzeboom, and Peter G.M. van der Heijden. 2003. "Patterns and Trends in Occupational Attainment of First Jobs in the Netherlands, 1930-1995: OLS Regression vs. Conditional Multinomial Logistic Regression. " Journal of the Royal Statistical Society 168 (part I):63- 84. DiPrete, Thomas A. 2002. "Life Course Risks, Mobility Regimes, and Mobility Consequences: A Comparison of Sweden, Germany, and the United States." American Journal of Sociology 108:267-309. DiPrete, Thomas A., Paul M. de Graaf, Ruud Luij kx, Michael Tahlin, and Hans-Peter Blossfeld. 1997. "Collectivist versus Individualist Mobility Regimes? Structural Change and Job Mobility in Four Countries." American Journal of Sociology 103 :3 1 8-358. DiPrete, Thomas A., and Patricia A. McManus. 1996. "Institutions, Technical Change, and Diverging Life Chances: Earnings Mobility in the United States and Germany." American Journal of Sociology 102:34-79. Driessen, Geert. 2004. "De taalsituatie van Caribische en mediterrane immigranten." Migrantenstudies 20:74-93. Duncan, Otis Dudley. 1966. "Path Analysis: Sociological. Examples." American Journal of Sociology 72: 1-16. —. 1968. "Inheritance of Poverty or Inheritance of Race?" Pp. 85-110 in On Understanding Poverty. Perspectives from the Social Sciences, edited by Daniel P. Moynihan. New York/London: Basic Books. Eitzen, D. Stanley, and Maxine Baca Zinn. 2004. Social Problems. Boston: Pearson Education. Engbersen, Godfried. 1999. "Panopticum Europa. Verblijfsstrategiee'n van illegale migranten." Pp. 219-241 in De ongekende stad I. Illegale vreemdelingen in Rotterdam, edited by Jack Burgers and Godfried Engbersen. Amsterdam: Boom. Entzinger, H.B., and P.J.J. Stij nen (Eds). 1990. Etnische Minderheden in Nederland. Meppel/Amsterdam: Boom/Ou. 262 Entzinger, Han. 2001. "Towards a Model of Incorporation: the Case of the Netherlands." Pp. 321-347 in Ethnic Minority and Interethnic Relations in Context. A Dutch-Hungarian Comparison, edited by Karen Phalet and Antal Orke'ny. Ashgate: Aldershot. Ericson, Robert, and John H. Goldthorpe. 1992a. "The CASMIN Project and the American Dream." European Sociological Review 81283-305. —. 1992b. The Constant Flux. A Study of C lass Mobility in Industrial Societies. Oxford: Clarendon Press. Esping—Andersen, Gosta. 1993. "Post-Industrial Class Structures: An Analytical Framework." Pp. 7-31 in Changing Classes. Stratification and Mobility in Post-Industrial Societies, edited by Gosta Esping-Andersen. London: SAGE. Esping-Anderson, Gosta. 1990. The Three Worlds of Welfare Capitalism. Princeton, NJ: Princeton University Press. Essed, Philomena. 1991. Understanding Everyday Racism. Newbury Park, CA: Sage Publications. Evans, Geoffrey. 1992. "Testing the Validity of the Goldthorpe class schema." European Sociological Review 8:211-232. Fainstein, Susan S. 1998. "Assimilation and Exclusion in US Cities." Pp. 28-44 in Urban Segregation and the Welfare State. Inequality and exclusion in western cities, edited by Sako Musterd and Wim Ostendorf. London and New York: Routledge. Faist, Thomas. 1995. Social Citizenship for Whom? Young Turks in Germany and Mexican Amerincans in the United States. Aldershot: Avebury. F alcon, Angelo. 2005. "De'tras Pa'lante: Explorations on the Future History of Puerto Ricans in New York City. " Pp. 147-192 in Boricuas in Gotham: Puerto Ricans in the Making of Modern New York City. Essays in the Memory of Antonia Pantoja, edited by Gabriel Haslip-Viera, Angelo F alcon, and Felix Matos Rodriguez. Princeton: Markus Wiener Publishers. F assmann, Heinz, and Rainer Milnz (Eds). 1994. European Migration in the Late 20th Century. Historical Patterns, Actual Trends, and Social Implications. Andershot, U.K.: E.Edgar. Flora, Peter, and Arnold J. Heidenheimer. 1981. "The Historical Core and Changing Boundaries of the Welfare State." in The Development of Welfare States in Europe and America, edited by Peter Flora and Arnold J. Heidenheimer. New Brunswick, NJ: Transaction Books. 263 F oner, Nancy. 2003. "Immi grants and African Americans: Comparative Perspectives on the New York Experience across Time and Space." Pp. 45-71 in Host Societies and the Reception of Immigrants, edited by Jeffrey G. Reitz. La Jolla, CA: Center for Comparative Immigration Studies. University of California, San Diego. Gans, Herbert J. 1968. "Culture and Class in the Study of Poverty: An Approach to Anto-Poverty Research. " Pp. 201-228 in On Understanding Poverty. Perspectives from the Social Sciences, edited by Daniel P. Moynihan. New York/London: Basic Books. Ganzeboom, Harry B. G., Ruud Luijkx, and Donald J. Treiman. 1989. "Intergenerational Class Mobility in Comparative Perspective." Reserach in Social Stratification and Mobility 8:3-84. Ganzeboom, Harry B.G., Donald J. Treiman, and Wout C. Ultee. 1991. "Comparative Intergenerational Stratification Research: Three Generations and Beyond." Annual Review of Sociology 17:277-302. Glazer, Nathan, and Daniel Patrick Moynihan. 1963. Beyond the Melting Pot. The Negroes, Puerto Ricans, Jews, Italians, and Irish of New York City. Cambridge: The MIT. Press. Gold, Steven J. 2004. "From Jim Crow to racial hegemony: Evolving explanations of racial hierarchy." Ethnic and Racial Studies 27:951-968. Goldthorpe, John H. 1980/1987. Social Mobility and Class Structure in Modern Britain. Oxford: Clarendon Press. Gowricharn, Ruben. 2004. "De duurzaamheid van het transnationalisme. De tweede generatie Hindostanen in Nederland." Migrantenstudies 20:252-2 86. Graaf, Paul M. de, and Ruud Luijkx. 1994. "Trends in Status Attainment in the Netherlands from Ascription to Achievement." in In Solidarity of Generations: Demographic, Economic, and Social Change, and its Consequences (Symposium). Amsterdam: Thesis Publishers. Gras, Mitzi, and Frank Bovenkerk. 1999. "Migrants and Ethnic Minorities in the Netherlands: Discrimination in Access to Employment. " Pp. 93-107 in Migrants, Ethnic Minorities and the Labour Market, edited by John Wrench, Andrea Rea, and Nouria Ouali. London: MacMillan Press. Grosfoguel, Ramon. 2003. Colonial Subjects. Puerto Ricans in a Global Perspective. Berkeley: University of California Press. 264 —. 2004. "La racializacion de 105 mi grantes coloniales del Caribe en los centros metropolitanos: Una introduccion a la historia de las diversas colonialidades en cada imperio." Caribbean Studies 32:3-42. Guzman, Pablo "Yoruba". 1980. "Puerto Rican Barrio Politics in the United States." Pp. 144-152 in Historical Perspectives on Puerto Rican Survival in the United States, edited by Clara Rodriguez and Virginia Sénchez Korrol. Princeton: Markus Wiener Publishers. Hagendoom, Louk. 1995. "Intergroup Biases in Multiple Group Systems: The Perception of Ethnic Hierarchies." Pp. 199-228 in European Review of Social Psychology, edited by Wolfgang Stroebe and Miles Hewstone: John Wiley & Sons Ltd. Hammar, Tomas. 1990. Democracy and the Nation State. Aliens, Denizens and Citizens in a World of International Migration. Aldershot: Avebury. Hamnet, Chris. 1998. "Social polarization, economic restructuring and welfare state regimes." Pp. 15-27 in Urban Segregation and the Welfare State. Inequality and exclusion in western cities, edited by Sako Musterd and Wim Ostendorf. London and New York: Routledge. Harris, J. R. , and Michael P. Todaro. 1970. "Migration, Unemployment and Development: A Two-Sector Analysis." American Economic Review 60: 139- 149. Haslip-Viera, Gabriel, and Sherrie L. Baver. 1996. "Introduction." Pp. 3-29 in Latinos in New York Communities in Transition, edited by Gabriel Haslip-Viera and Sherrie L. Baver. Notre Dame/London: University of Notre Dame Press. Heelsum, A]. van. 1997. De etnisch-culturele positie van de tweede generatie Surinamers. Amsterdam: Het Spinhuis. Hernandez, J 038'. 1996. "The Identity and Culture of Latino College Students." Pp. 126-146 in Latinos in New York. Communities in Transition, edited by Gabriel Haslip—Viera and Sherrie L. Baver. Notre Dame/London: Notre Dame University Press. Hiebert, Dan. 2002. "Canadian immigration and the selection-settlement services trade-off: Exploring immigrant economic participation in British Columbia." edited by RIIM. Vancouver. Hiebert, Dan, Jock Collins, and Paul Spoonley. 2003. "Uneven Globalization: Neoliberal Regimes, Immigration, and Multiculturalism in Australia, Canada, and New Zealand." in Working Paper Series No. 03-05: Research on Immigration and Integration in the Metropolis, edited by RIIM. Vancouver. 265 Hinojosa-Ojeda, Raul, Martin Camoy, and Hugh Daley. 1991. "An Even Greater "U- Turn". Latinos and the New Inequality." Pp. 25-52 in Hispanics in the Labor Force. Issues and Politics, edited by Edwin Melendez, Clara E. Rodriguez, and Janis Barry Figueroa. New York: Plenum Press. Hirschman, Charles. 1988. "Minorities in the Labor Market. Cyclical Patterns and Secular Trends in J oblessness." Pp. 63-85 in Divided Opportunities. Minorities, Poverty, and Social Policy, edited by Gary D. Sandefur and Marta Tienda. New York/London: Plenum Press. Hirschman, Charles, and Morrison G. Wong. 1984. "Socioeconomic Gains of Asian Americans, Blacks, and Hispanics: 1960-1976." American Journal of Sociology 90:5 84-607. Hobsbawm, E.J. 1990/1992. Natie en nationalisme sedert I780. Streven, mythe en werkelijkheid. Amsterdam: Uitgeverij Jan Mets. Hoerder, Dirk, and Leslie Page Moch. 1996. European Migrants. Global and Local Perspectives. Boston: Northeastern University Press. Hollifield, James F. 1992. Immigrants, Markets, and States. The Political Economy of Postwar Europe. Cambridge MA/London: Harvard University Press. Howell, David R., and Elizabeth J. Mueller. 1997. "The Effects of Immigrants on African-American Earnings: A J obs-Level Analysis of the New York City Labor Market, 1979-89." New York: Working Paper 210; Robert J. Milano Graduate School. Hulst, Hans van. 2000. "A continuing construction of crisis. Antilleans, especially Curacaoans, in the Netherlands." Pp. 93-122 in Immigrant Integration. The Dutch Case, edited by Hans Vermeulen and Rinus Penninx. Amsterdam: Het Spinhuis. J anoski, Thomas. 1998. Citizenship and Civil Society. A Framework of Rights & Obligations in Liberal, Traditional, and Social Democratic Regimes. Cambridge: Cambridge University Press. Jensen, Leif. 1989. The New Immigration. Implicationsfor Poverty and Public Assistance Utilization. New York: Greenwood Press. Jonsson, Jan 0., and Colin Mills. 1993. "Social Mobility in the 19705 and 19805: A Study of Men and Women in England and Sweden." European Sociological Review 9229-248. 266 J oppke, Christian. 1998a. "Asylum and State Sovereignty: A Comparison of the United States, Germany, and Britain." Pp. 109-152 in Challenge to the Nation-State. Immigration in Western Europe and the United States, edited by Christian Joppke: Oxford University Press. -— (Ed). 1998b. Challenge to the Nation-State. Immigration in Western Europe and the United States. Oxford: Oxford University Press. —. 1998c. "Immigration Challenges the Nation-State." Pp. 5-46 in Challenge to the Nation-State. Immigration in Western Europe and the United States, edited by Christian J oppke: Oxford University Press. Kasinitz, Philip, John Mollenkopf, and Mary C. Waters. 2003. "Becoming Americans/Becoming New Yorkers: hnmi grant Incorporation in a Majority Minority City." Pp. 73-90 in Host Societies and the Reception of Immigrants, edited by Jeffrey G. Reitz. La Jolla, CA: Center for Comparative Immigration Studies University of California, San Diego. Kerckhoff, Alan C. 1995. "Institutional Arrangements and Stratification Processes in Industrial Societies." Annual Review of Sociology 15:323-47. Kersbergen, Kees van. 2000. "A caring state or a could-not care-less state?" The Netherlands'Journal of Social Sciences 36:48-51. Kim, J ea-On, and Charles W. Mueller. 1978. Introduction to Factor Analysis. What It Is and How To Do It. Newbury Park: SAGE Publications. Klerk, Leo de, and Hans van Amersfoort. 1988. "Surinamese settlement in Amsterdam 1973-83." Pp. 147-163 in Lost Illusions. Caribbean minorities in Britain and the Netherlands, edited by Malcolm Cross and Han Entzinger. London: Routledge. Kloosterman, RC. 1991. "Stedelijke arbeidsmarktparadoxen: Het Amsterdamse voorbeeld." Pp. 25-41 in De werkende stad: Aspecten van de grootstedelijke arbeidsmarkt, edited by O.A.L.C. Atzema, M. Hessels, and H. Zondag. Utrecht: Stedelijke Netwerken, Elinkwijk bv. Kloosterman, Robert C. 1994. "Three Worlds of Welfare Capitalism? The Welfare State and the Post-Industrial Trajectory in the Netherlands after 1980." West European Politics 17: 166- 1 89. Kloosterman, Robert, Joanne van der Leun, and Jan Rath. 1998. "Across the border: immigrants' economic opportunities, social capital and informal business activities." Journal of Ethnic and Migration Studies 24:249-268. 267 —. 1999. "Mixed Embeddedness: (In)formal Economic Activities and Immigrant Businesses in the Netherlands." submitted for International Journal of Urban and Regional Research. Koot, Willem, and Petrien Uniken Venema. 1988. "Education: the way up for Surinamese in the Netherlands?" Pp. 185-203 in Lost Illusions. Caribbean minorities in Britain and the Netherlands, edited by Malcolm Cross and Han Entzinger. London: Routledge. Korteweg, Anna C. 2003. "The Construction of the Masculine Citizen-Worker: An Ethnographic Comparison of Welfare-to-Work Workshops in the Netherlands and the United States." in RC 19. Toronto. —. 2005. "Between Punishment and Well-Being: The Role of Ideas in the Work of Welfare in the United States and the Netherlands." in Annual Conference of ISA Research Committee 19: Poverty, Social Welfare, and Social Policy. Chicago. Korzeniewicz, Roberto Patricio, Timothy P. Moran, and Angela Stach. 2003. "Trends in Inequality: Toward a World-Systems Analysis." in Globalization and Society. Processes of Differentiation Examined, edited by Raymond Breton and Jeffrey G. Reitz. Westport, CT: Praeger. LaLonde, Robert J ., and Robert H. Topel. 1992. "The Assimilation of Immigrants in the US. Labor Market." Pp. 67-92 in Immigration and the Work F orce. Economic Consequences for the United States and Source Areas, edited by George J. Borj as and Richard B. Freeman. Chicago and London: The University of Chicago Press. Landale, Nancy S., and RS. Oropesa. 2002. "White, Black, or Puerto Rican? Racial Self-Identification among Mainland and Island Puerto Ricans." Social Forces 81 :23 1-254. Lee, Sharon M. 1993. "Racial Classifications in the US. Census: 1890-1990." Ethnic and Racial Studies 16:75-94. Lentin, Alana. 2000. "'Race', Racism, and Anti-racism: Challenging Contemporary Classifications." Social Identities 6291-106. Lewin-Epstein, Noah, and Moshe Semyonov. 1986. "Ethnic Group Mobility in the Israeli Labor Market." American Sociological Review 5 1 :342-3 52. Lewis, Oscar. 1965. La Vida. A Puerto Rican Family in the Culture of Poverty - San Juan and New York. New York: Random House. 268 —. 1968. "The Culture of Poverty." Pp. 187-200 in On Understanding Poverty. Perspectives from the Social Sciences, edited by Daniel P. Moynihan. New York/London: Basic Books. Light, Ivan, Parminder Bhachu, and Stavros Karageorgis. 1993. "Migration Networks and ImmigrantEntrepreneurship." Pp. 25-49 in Immigration and Entrepreneurship. Culture, Capital, and Ethnic Networks, edited by Ivan Light and Parminder Bhachu. New Brunswick: Transaction Publishers. Light, Ivan, and Steven J. Gold. 2000. Ethnic Economies. San Diego: Academic Press. Lindeman, Ellen, Kees Dignum, Peggy Schyns, Karen Klein Wolt, Harry Smeets, Simone Crok, and Jeroen Slot. 2003. "De Staat van de Stad Amsterdam H. Ontwikkelingen in participatie en leefsituatie." Amsterdam: Gemeente Amsterdam. Dienst Onderzoek en Statistiek. 3“ “Tm“:l Lipset, Seymour Martin , and Reinhard Bendix. 1967. Social Mobility in Industrial Society. Berkeley and Los Angeles: University of California Press. Livi-Bacci, M. 1993. "South-North Migration: A Comparative Approach to North American and European Experiences." in OECD. Paris. Maldonaldo, E. 1979. "Contract Labor and the Origins of Puerto Rican Communities in the US." International Migration Review 13:103-121. Marger, Martin N. 2003. Race and Ethnic Relations. American and Global Perspectives. Belmont, CA: Thomson/Wadsworth. Marshall, TH. 1965. Class, Citizenship, and Social Development. Garden City, NY.: Anchor Books. Martens, BF, and A0. Verweij. 1997. Surinamers in Nederland Kerncyfers I 996. Rotterdam: Iinstituut voor Sociologisch-Economisch Onderzoek. Martin, Philip L. 1994. "The United States: Benign Neglect toward Immigration." Pp. 83-99 in Controlling Immigration. A Global Perspective, edited by Wayne A. Cornelius, Philip L. Martin, and James F. Hollifield. Stanford, CA: Stanford University Press. Massey, Douglas. 1990. "American Apartheid: Segregation and the Making of the Underclass. " American Journal of Sociology 96:329-357. Massey, Douglas, Joaquin Arango, Graeme Hugo, Ali Kouaouci, Adela Pellegrino, and J. Edward Taylor. 1993. "Theories of lntemational Migration: A Review and Appraisal." Population and Development Review 19:431-466. 269 —. 1994. "An Evaluation of lntemational Migration Theory: The North American Case." Population and Development Review 20:699-751. Massey, Douglas S. 1999. "International Migration at the Dawn of the Twenty-First Century: The Role of the State." Population and Development Review 25:303- 322. Massey, Douglas S., Joaquin Arango, Graeme Hugo, Ali Kouaouci, Adela Pellegrino, and J. Edward Taylor. 1998. Worlds in Motion. Understanding International Migration at the End of the Millenium. Oxford: Clarendon Press. Massey, Douglas S., and Nancy A. Denton. 1989. "Residential Segregation of Mexicans, Puerto Ricans, and Cubans in Selected U.S. Metropolitan Areas." Social Science Review 73:73-83. —. 1993. American Apartheid. Segregation and the Making of the Underclass. Cambridge: Harvard University Press. Matras, Judah. 1980. "Comparative Social Mobility." Annual Review of Sociology 6:401-431. Melendez, Edwin. 1991. "Labor Market Structure and Wage Differences in New York City. A Comparative Analysis of Hispanics and Non-Hispanic Blacks and Whites." Pp. 101-1 18 in Hispanics in the Labor Force. Issues and Policies, edited by Edwin Melendez, Clara E. Rodriguez, and Janis Barry Figueroa. New York: Plenum Press. Melendez, Edwin, Clara E. Rodriguez, and Janis Barry Figueroa. 1991. "Hispanics in the Labor Force: And Introduction to Issues and Approaches." Pp. 1-21 in Hispanics in the Labor F orce. Issues and Politics. , edited by Edwin Melendez, Clara E. Rodriguez, and Janis Barry Figueroa. New York: Plenum Press. Menjivar, Cecilia. 2000. Fragmented Ties. Salvadorian Immigrant Networks in America. Berkeley/Los Angeles/London: University of California Press. Miller, Robert L. 2001. "The Industrial Context of Occupational Mobility: Change in Structure." Research in Social Stratification and Mobility 18:313-353. Miller, SM. 1960. "Comparative Social Mobility." Current Sociology-La Sociologie Contemporaine 9: 1-89. Minderhoud, Paul. 1998. "Regulation of Migration: Introduction." Pp. 7-24 in Regulation of Migration. International Experiences, edited by Anita BOCker, Kees Groenendijk, Tetty Havinga, and Paul Minderhoud. Amsterdam: Het Spinhuis Publishers. 270 Moch, Leslie Page. 1992. Moving Europeans: Migration in Western Europe Since 1650. Bloomington & Indianapolis: Indiana University Press. Model, Suzanne, and David Lapido. 1996. "Context and Opportunity: Minorities in London and New York. " Social Forces 75:485-510. Mollenkopf, John. 2000. "Assimilating immigrants in Amsterdam: A perspective from New York." The Netherlands'Journal of Social Sciences 36:15-34. Morales, Julio. 1986. Puerto Rican Poverty and Migration. We just had to try elsewhere. New York: Praeger. Moss, Philip, and Chris Tilly. 1995. ’Soft Skills'and Race: An Investigation of Black Men's Employment Problems. New York: Russell Sage Foundation. Mullan, Brendan. 1998. "The Regulation of lntemational Migration: The US and Western Europe in Historical Comparative Perspective." Pp. 27-44 in Regulation of Migration. International Experiences, edited by Anita BOcker, Kees Groenendijk, Tetty Havinga, and Paul Minderhoud. Amsterdam: Het Spinhuis Publishers. M ullan, Brendan, and Lorraine Majka. 1997. "The Utilization of Public Assistance Among Refugees in the Chicago Metropolitan Area." Munz, R. 1995. "Where did they all come from? Typology and geography of European mass migration in the twentieth century." in Plenary Session 111,- European Population Conference. Milano, Italy. Musterd, Sako. 2001. "Immigration and Ethnic Segregation in the Netherlands with a Special Focus on Amsterdam." in Ethnic Minorin and Interethnic Relations in Context. A Dutch-Hungarian Comparison, edited by Karen Phalet and Antal Orkeny. Ashgate: Aldershot. Musterd, Sako, and Wim Ostendorf. 1993. "Stedelijke arrnoede en etniciteit in de verzorgingsstaat; Amsterdam als voorbeeld." Sociologische Gids XL:466-48l. Musterd, Sako, Wim Ostendorf, and Matthijs Breebaart. 1998. Multi-Ethnic Metropolis: Patterns and Policies. Dordrecht: Kluwer Academic Publishers. Myrdal, Gunnar. 1960. Beyond the Welfare State. New Haven Connecticut: Bantam Book/Yale University Press. Nakao, Keiko. 1992. "Occupations and Stratification: Issues of Measurement." Contemporary Sociology 21 :65 8-662. 271 Niekerk, Mies van. 2000a. De Krekel en de Mier. Fabels enfeiten over maatschappelijke steiging van Creoolse en H indoestaanse Surinamers in Nederland. Amsterdam: Het Spinhuis. —. 2000b. "Paradoxes in paradise. Integration and social mobility of the Surinamese in the Netherlands." Pp. 64-92 in Immigrant Integration. The Dutch Case, edited by Hans Vermeulen and Rinus Penninx. Amsterdam: Het Spinhuis. —. 2004. "Afro-Caribbeans and Indo-Carribeans in the Netherlands: Premigration Legacies and Social Mobility." International Migration Review 38:158-183. North, David S. 1978. Seven Years Later: The Experiences of the 1970 Cohort of Immigrants in the US. Labor Market. Washington DC: Linton & Company, Inc. O'Conner, James. 1973. The Fiscal Crisis of the State. New York: St. Martin's Press. ——-—. 2001. The Fiscal Crisis of the State, 2nd Ed. (Introduction). Somerset, NJ: Transaction Publishers. O+S, Het Amsterdamse Bureau voor Onderzoek en Statistiek. 1991. "Jaarboek Amsterdam in Cij fers 1991." edited by Wim Van Zee, Ton Brinkman, and Windsant Vermeulen, Jos: Stadsdrukkerij Amsterdam. —. 2002. "Jaarboek Amsterdam in Cijfers 2002." edited by Wim van Zee, Cathelijne Hermans, Cor Hylkema, and E15 de Boer. Amsterdam: Stadsdrukkerij Amsterdam. Omi, Michael, and Howard Winant. 1986/1989. Racial Formation in the United States: From the I 960s to the I980s. New York: Routledge. Oostindie, Gert J. 1988. "Caribbean migration to the Netherlands: a journey to disappointment?" Pp. 54-72 in Lost Illusions. Caribbean minorities in Britain and the Netherlands, edited by M. Cross and H. Entzinger. London: Routledge. Ortiz, Vilma. 1986. "Changes in the Characteristics of Puerto Rican Migrants from 1955 to 1980." International Migration Review 20:612-628. —. 1991. "Latinos and Industrial Change in New York and Los Angeles." Pp. 119- 132 in Hispanics in the Labor Force. Issues and Policies, edited by Edwin Melendez, Clara E. Rodriguez, and Janis Barry Figueroa. New York: Plenum Press. Ottaviano, Gianmarco IR, and Giovanni Peri. 2005. "Rethinking the Gains from Immigration: Theory and Evidence from the US." NBER Working Papers: W11672. 272 Pedraza, Silvia, and Ruben G. Rumbaut. 1996. Origins and Destinies Immigration, Race, and Ethnicity in America. Belmont, CA: Wadworth Publishing Company. Penninx, Rinus. 2005. "Bridges between Research and Policy? The Case of Post-War Immigration and Integration Policies in the Netherlands." Pp. 33-48 in International Journal on Multicultural Societies: www.unesco.orglshsflims/von/issuel/art3. Pe'rez y Gonzalez, Maria E. 2000. Puerto Ricans in the United States. Westport: Greenwood Press. Petti grew, Thomas F. 1998. "Reactions Towards the New Minorities of Western Europe." Annual Review of Sociology 24:77-103. Polanyi, Karl. 1944. The Great Transformation. The Political and Economic Origins of Our Time. Boston: Beacon Press. Portes, Alejandro, William Haller, and Luis E. Guamizo. 2001. "Transnational Entrepreneurs: The Emergence and Determinants of an Alternative Form of Immigrant Economic Adaptation." in WPT C . Portes, Alejandro, and Leif Jensen. 1989. "The Enclave and the Entrants: Patterns of Ethnic Enterprise in Miami before and after Mariel." American Sociological Review 54:929-949. Portes, Alejandro, and Rube'n Rumbaut. 1996. Immigrant America. A Portrait. Berkeley/Los Angeles/London: University of California Press. Portes, Alejandro, and Saskia Sassen-Koob. 1987. "Making it Underground: Comparative Material on the Informal Sector in Western Market Economies.‘ American Journal of Sociology 93130-6 l. f Portes, Alejandro, and Min Zhou. 1996. "Self-Employment and the Earnings of Immigrants." American Sociological Review 61:219-230. Praag, Carlo van. 2003. "Demografie." in Rapportage Minderheden 2003. Onderwijs, arbeid en sociaal-culturele integratie, edited by J aco Dagevos, Mérove Gij sberts, and Carlo van Praag. Den Haag: Sociaal en C‘ultureel Planbureau. Raes, Stephan. 2000. "Regionalisation in a Globalising World: The Emergence of Clothing Sweatshops in the European Union." Pp. 20-36 in Immigrant Businesses: The Economic, Political and Social Environment. , edited by Jan Rath. London/New York: MacMillan Press/St. Martin's Press. 273 Raes, Stephan, Jan Rath, Marja Dreef, Adem Kumcu, Flavia Reil, and Aslan Zorlu. 2002. "Amsterdam: Stiched up." Pp. 89-111 in Unravelling the Rag Trade: Immigrant Entrepreneurship in Seven World Cities, edited by Jan Rath. Rath, Jan. 1999. "The Netherlands: A Dutch Treat for Anti-Social Families and Immigrant Ethnic Minorities." in The European Union and Migrant Labour, edited by Gareth Dale and Mike Cole. Oxford/New York: Berg. —. 2000. "Assimilation of immigrants in Amsterdam and New York: A case for international comparison." The Netherlands'Journal of Social Sciences 36:5- 14. —. 20028. "Needle Games: A Discussion of Mixed Embeddedness." Pp. 1-27 in Unravelling the Rag Trade. Immigrant Entrepreneurship in Seven World Cities, edited by Jan Rath. —. 2002b. "Sewing up Seven Cities." Pp. 169-191 in Unravelling the Rag Trade. Immigrant Entrepreneurship in Seven World Cities, edited by Jan Rath. Amsterdam: Het Spinhuis. .Razin, Eran. 1993. "Immigrant Entrepreneurs in Israel, Canada, and California." Pp. 97-124 in Immigration and Entrepreneurship. Culture, Capital, and Ethnic Networks, edited by Ivan Light and Parminder Bhachu. New Brunswick/London: Transaction Publishers. Rea, Andrea, John Wrench, and Nouria Ouali. 1999. "Introduction: Discrimination and Diversity." Pp. 1-18 in Migrants, Ethnic Minorities and the Labour Market, edited by John Wrench, Andrea Rea, and Nouria Ouali. London: MacMillan Press. Reimers, Cordelia W. 1985. "A Comparative Analysis of the Wages of Hispanics, Blacks, and Non-Hispanic Whites." Pp. 27-75 in Hispanics in the US. Economy, edited by George J. Borjas and Marta Tienda. Orlando: Academic Press, Inc. Reimers, Cornelia, and Howard Chemick. 1991. "Hispanic Employment in the Public Sector. Why Is It Lower Than Blacks'?" Pp. 135-157 in Hispanics in the Labor Force. Issues and Policies, edited by Edwin Melendez, Clara E. Rodriguez, and Janis Barry Figueroa. New York: Plenum Press. Reitz, Jeffrey G. 1998. Warmth of Welcome. The Social Causes of Economic Success for Immigrants in Diflerent Nations and Cities. Boulder: WestviewPress. Reubsaet, T.J.M. 1982. Surinaamse migranten in Nederland. Het beleid van de rijksoverheid van I 965-I979: uitgangspunten, doelstellingen en middelen. Nijmegen: Instituut voor Toegepaste Sociologie. 274 Reubsaet, T.J.M., and J .A. Kropman. 1983. Surinaamse migranten in Nederland. De positie van Surinamers en het gevoerde overheidsbeleid Nij megen: Instituut voor Toegepaste Sociologie. Reubsaet, T.J.M., J .A. Kropman, and L.M. van Mulier. 1982. Surinaamse migranten in Nederland. De positie van Surinamers in de Nederlandse samenleving. Nijmegen: Instituut voor Toegepaste Sociologie. Reubsaet, Theo. 1988. "On the way up? Surinamese and Antilleans in the Dutch labour market." in Lost Illusions. Caribbean minorities in Britain and the Netherlands, edited by Malcolm Cross and Han Entzinger. London: Routledge. Rivera-Batiz, Francisco L. 1991. "The Effects of Literacy on the Earnings of Hispanics in the United States." Pp. 53-75 in Hispanics in the Labor force. Issues and Policies, edited by Edwin Melendez, Clara E. Rodriguez, and Janis Barry Figueroa. New York: Plenum Press. —. 2005. "Puerto Rican New Yorkers in the 19905: A Demographic and Socioeconomic Profile." in Boricuas in Gotham: Puerto Ricans in the Making of Modern New York City. Essays in Memory of Antonia Pantoja, edited by Gabriel Haslip—Viera, Angelo F alcon, and Felix Matos Rodriguez. Princeton: Markus Wiener Publishers. Rodriguez, Clara. 1980a. "Economic Survival in New York City." Pp. 38-53 in Historical Perspectives in Puerto Rican Survival in the United States, edited by Clara E. Rodriguez and Virginia Sénchez Korrol. Princeton: Markus Wiener Publications. Rodriguez, Clara E. 1980b. "Puerto Ricans: Between Black and White." Pp. 25-35 in Historical Perspectives on Puerto Rican Survival in the United States, edited by Clara E. Rodriguez and Virginia Sénchez Korrol. Princeton: Markus Wiener Publications. —. 1989. Puerto Ricans. Born in the USA. Boston: Unwin Hyman. —. 1991. "The Effect of Race on Puerto Rican Wages." Pp. 77-98 in Hispanics in the Labor Force. Issues and Policies, edited by Edwin Melendez, Clara E. Rodriguez, and Janis Barry Figueroa. New York: Plenum Press. —. 1996. "Racial Themes in the Literature: Puerto Ricans and Other Latinos." Pp. 104-125 in Latinos in New York. Communities in Transition, edited by Gabriel Haslip-Viera and Sherrie L. Baver. Notre Dame/London: Notre Dame University Press. 275 Rodriguez, Clara, Virginia sanchez Korrol, and Jose Oscar Alers. 1980. "The Puerto Rican Struggle to Survive in the United States." Pp. 1-10 in Historical Perspectives on Puerto Rican Survival in the (1.5., edited by Clara E. Rodriguez and Virginia Sénchez Korrol. Princeton: Marcus Wiener Publishers. Roelandt, Theo, and Justus Veenman. 1994. "Ethnic groups and underclass formation in the Netherlands: The Dutch ethno-class structure." in Immigrant Ethnic Minorities in the Dutch Labour Market. Analyses and Policies, edited by Han Entzinger, Jacques Siegers, and F rits Tazelaar. Amsterdam: Thesis Publishers. Rose, Joseph B. 1994. Puerto Rican New Yorkers. New York: Department of City Planning. Rumbaut, Ruben. 2000. "Defining the Situation: The Ethnic Identities of Children of Immigrants." in Legacies. The Story of the New Second Generation, edited by Alejandro Portes and Ruben Rumbaut. Berkeley/New York: University of California Press/Russell Sage Publications. Rumbaut, Ruben G. 1996. "The Crucible Within: Ethnic Identity, Self-Esteem, and Segmented Assimilation Among Children of Immigrants." Pp. 119—170 in The New Second Generation, edited by Alejandro Portes. New York: Russell Sage Foundation. sanchez Korrol, Virginia. 1980. "Survival of Puerto Rican Women in New York before World War II." Pp. 57-67 in Historical Perspectives on Puerto Rican Survival in the United States, edited by Clara Rodriguez and Virginia Sénchez Korrol. Princeton: Markus Wiener Publishers. —. 2005. "Building the New York Puerto Rican Community, 1945-1965: A Historical Interpretation." Pp. 1-18 in Boricuas in Gotham: Puerto Ricans in the Making of Modern New York City. Essays in the Memory of Antonia Panto/a, edited by Gabriel Haslip—Viera, Angelo Falcon, and Felix Matos Rodriguez. Princeton: Markus Wiener Publishers. Sandefur, Gary D., and Marta Tienda. 1988. "Introduction. Social Policy and the Minority Experience." Pp. 1-19 in Divided Opportunities. Minorities, Poverty, and Social Policy, edited by Gary D. Sandefur and Marta Tienda. New York/London: Plenum Press. Sansone, Livio. 1990. Last Boto. De boot gemist. Over Surinaamsejongeren, werk en werk/oosheid. Amersfoort/Leuven: Acco. 276 —. 1992. Schitteren in de schaduw. Overlevingsstrategiee'n, subcultuur en etniciteit van Creoolsejongeren uit de lagere klasse in Amsterdam 1 981-1990. Amsterdam: Het Spinhuis. Santana Cooney, Rosemary, and Alice Colon. 1980. "Work and Family: The Recent Struggle of Puerto Rican Females." Pp. 70-85 in Historical Perspectives on Puerto Rican Survival in the United States, edited by Clara E. Rodriguez and Virginia Sénchez Korrol. Princeton: Markus Wiener Publishers. Sassen-Koob, Saskia. 1985. "Changing Composition and Labor Market Location of Hispanic Immigrants in New York City, 1960-1980." Pp. 299-322 in Hispanics in the US. Economy, edited by George J. Borjas and Marta Tienda. Orlando: Academic Press, Inc. Sassen, Saskia. 1988. The Mobility of Labor and Capital: A Study in International Investment and Labor F low. Cambridge: Cambridge University Press. —. 1995a. "Immigration and Local Labor Markets." in The Economic Sociology of Immigration. Essays on Networks, Ethnicity, and Entrepreneurship, edited by Alejandro Portes. New York: Russell Sage Foundation. —. 1995b. Losing Control? Sovereignty in an Age of Globalization. New York: Columbia University Press. —. 2000. Cities in a World Economy. Thousand Oaks: Pine Forge Press. Schoeni, Robert F., Kevin F. McCarthy, and Georges Vemez. 1996. The Mixed Economic Progress of Immigrants. Santa Monica, CA: Rand. Schoorl, Jeanette, Liesbeth Heering, Ingrid Esveldt, George Groenewold, Rob van der Erf, Alinda Bosch, Helga de Valk, and Bart de Bruijn. 2000. "The Push and Pull Factors of lntemational Migration. A Comparative Report. " in Theme 1, General Statistics: European Commission/Eurostat. Schuster, John. 1999. Poortwachters over immigranten. Het debat over immigratie in het naoorlogse Groot-Brittannie' en Nederland. Amsterdam: Het Spinhuis. Smith, Anthony D. 1991. National Identity. London: Penguin Books. Smith, James P. 1988. "Poverty and the Family." Pp. 141-172 in Divided Opportunities. Minorities, Poverty, and Social Policy, edited by Gary D. Sandefur and Marta Tienda. New York/London: Plenum Press. 277 Smith, Robert C., Hector R. Cordero-Guzman, Ramon Grosfoguel. 2001. "Introduction: Migration, Transnationalization, and Ethnic and Racial Dynamics in a Changing New York." in Migration, T ransnationalization, and Race in a Changing New York, edited by Robert C. Smith Hector R. Cordero- Guzman, Ramon Grosfoguel. Philadelphia: Temple University Press. Stafford, Walter. 1991. "Racial, Ethnic, and Gender Employment Segmentation in New York City Agencies." Pp. 159-180 in Hispanics in the Labor Force. Issues and Policies, edited by Edwin Melendez, Clara E. Rodriguez, and Janis Barry Figueroa. New York: Plenum Press. Stahl, Charles W. 1995. "Theories of lntemational Labor Migration: An Overview." The Asian and Pacific Migration Journal 4:21 1-232. Staring, Richard. 1999. "Migratiescenario's. De overkomst van illegale migranten." Pp. 54-87 in De ongekende stad I. Illegale vreemdelingen in Rotterdam, edited by Jack Burgers and Godfried Engbersen. Amsterdam: Boom. Sturz, Herbert. 1985. "The Puerto Rican New Yorkers: Part H Socioeconomic Characteristics and Trends 1970-1980." New York: Department of City Planning. Swaan, Abram de. 1996. Zorg en de staat. Welzijn, onderwifs en gezondheidszorg in Europa en de Verenigde Staten in de nieuwe tijd. Amsterdam: Bert Bakker. Tesser, P.T.M. 1993. Rapportage Minderheden I 993. Den Haag: Sociaal en Cultureel Planbureau. Thompson, Bruce. 2004. Exploratory and Confirmatory Factor Analysis. Understanding Concepts and Applications. Washington, DC: American Psychological Association. Tienda, Marta. 1993. "Puerto Ricans and the Underclass Debate." Pp. 122-136 in The Ghetto Underclass. Social Science Perspectives, edited by William Julius Wilson. Newbury Park: SAGE Publications. Tienda, Marta, and Patricia Guhleman. 1985. "The Occupational Position of Employed Hispanic Women " Pp. 243-273 in Hispanics in the US. Economy, edited by George J. Borjas and Marta Tienda. Orlando: Academic Press, Inc. Tjong Kim Sang, Erik. 1995 Torres, Andres, and Clara E. Rodriguez. 1991. "Latino Research and Policy. The Puerto Rico Case." Pp. 247-263 in Hispanics in the Labor Force. Issues and Policies, edited by Edwin Melendez, Clara E. Rodriguez, and Janis Barry Figueroa. New York: Plenum Press. 278 Torres, Rodolfo D., and Adela de la Torre. 1991. "Latinos, Class, and the US. Political Economy. Income Inequality and Policy Alternatives." Pp. 265-287 in Hispanics in the Labor Force. Issues and Policies, edited by Edwin Melendez, Clara E. Rodriguez, and Janis Barry Figueroa. New York: Plenum Press. Treiman, Donald J. 1977. Occupational Prestige in Comparative Perspective. New York: Academic Press. —. 2001. "Occupations, Stratification, and Mobility." Pp. 298-313 in The Blackwell Companion to Sociology, edited by Judith R. Blau: Blackwell. Treiman, Donald J ., and Harry B. G. Ganzeboom. 2000. "The Fourth Generation of Comparative Stratification Research." Pp. 123-150 in The International Book of Sociology, edited by Stella R. Quak and Amaud Sales. London: Sage. Treiman, Donald J ., and Kam-Bor Yip. 1989. "Educational and Occupational Attainment in 21 Countries." Pp. 373-394 in Cross-National Research in Sociology, edited by Melvin L. Kohn. Beverly Hills: Sage Publications. Tubergen, Frank van, Ineke Maas, and Henk Flap. 2004. "The Economic Incorporation of Immigrants in 18 Western Societies: Origin, Destination, and Community Effects." American Sociological Review 69:704-727. Turner, Bryan S. 2006. "Global Society and the Nature of Rights." Societies without Borders 1:41-52. Turner, J. Rick, and Julian F. Thayer. 2001. Introduction to the Analysis of Variance. Design, Analysis, and Interpretation. Thousand Oaks: Sage Publications. Ultee, Wout, and Ruud Luijkx. 1986. "Intergenerational Standard-of-Living Mobility in Nine EEC Countries: Country Characteristics, Competitive Balance and Social Fluidity." European Sociological Review 2: 191-207. Valkonet-Freeman, Marielle. 1978/1979. "De 'gesloten buurten' van Amsterdam." Pp. 58-77 in Omdat zij anders zijn. Patronen van rasdiscriminatie in Nederland, edited by Frank Bovenkerk. Meppel/Amsterdam: Boom. Vasta, Ellie. 2006. "From Ethnic Minorities to Ethnic Majority Policy: changing identities and the shift to assimilationism in the Netherlands." COMPAS, University of Oxford. Draft. Veenman, Justus. 2001. "Ethnic Minority Educational and Labour Market Performance in the Netherlands." Pp. 231-247 in Ethnic Minority and Interethnic Relations in Context. A Dutch-Hungarian Comparison, edited by Karen Phalet and Antal Orkény. Ashgate: Aldershot. 279 Veenman, Justus, and Theo Roelandt. 1994. "Ethnic minorities in the Dutch labor market: an overview of facts and theories." Pp. 25-45 in Immigrant Ethnic Minorities in the Dutch Labour Market. Analyses and Policies, edited by Han Entzinger, Jacques Siegers, and Frits Tazelaar. Amsterdam: Thesis Publishers. Venema, T. 1992. Famiri nanga kulturu. Creoolse sociale verhoudingen en Winti in Amsterdam. Amsterdam: Het Spinhuis. Vermeulen, Hans, and Rinus Penninx. 1994. "Inleiding." Pp. 1-14 in Het democratisch ongeduld. De emancipatie en integratie van zes doelgroepen van het minderhedenbeleid. , edited by Hans Vermeulen and Rinus Penninx. Amsterdam: Het Spinhuis. —. 2000. "Introduction." Pp. 1-35 in Immigrant Integration. The Dutch Case, edited by Hans Vermeulen and Rinus Penninx. Amsterdam: Het Spinhuis. Visser, Jelle, and Anton Hemerijck. 1997. A Dutch Miracle. Job Growth, Welfare reform and C orporatism in the Netherlands. Amsterdam: Amsterdam University Press. Wade, Peter. 1997. Race and Ethnicity in Latin America. London/Chicago: Pluto Press. Waldinger, Roger. 1989. "Structural Opportunity or Ethnic Advantage? Immigrant Business Development in New York. " International Migration Review 23:48- 72. —-—-. 2001. "Strangers at the Gates." Pp. 1-29 in Strangers at the Gates. New Immigrants in Urban America, edited by Roger Waldinger. Berkeley: University of California Press. Wallerstein, Immanuel. 1974. The Modern World System 1. San Diego: Academic Press. —. 1980. The Modern World-System II. Mercantilism and the Consolidation of the European World Economy 1600-1 750. New York: Academic Press. —. 1995. "The Modern World-System and Evolution." Journal of World-Systems Research 1. Waters, Mary. 1996. "Ethnic and Racial Identities of Second-Generation Black Immigrants in New York City." Pp. 171-196 in The New Second Generation, edited by Alejandro Portes. New York: Russell Sage Foundation. 280 Weiss, David J. 2006. Analysis of Variance and Functional Measurement. Oxford: Oxford University Press. Wilders, Geert. 2005. "Den Haag is laf tegen islamistisch extremisme. Beantwoord haat en geweld van terroristen met uitsluiting en intolerantie. (The Hague responds cowardly towards islamic extremism. The answer to hate and violence is exclusion and intolerance)" Pp. 12 in NRC/Handelsblad Web- en Weekeditie voor het buitenland. Rotterdam. Wilensky, Harold L., and Charles N. Lebeaux. 19658. Industrial Society and Social Welfare. The impact of industrialization on the supply and organization of welfare services in the United States. New York: The Free Press. —. 1965b. "Introduction." Pp. 1-10 in Social Welfare Institutions. A Sociological Reader, edited by Mayer N. Zald. New York: John Wiley & Sons. Wilson, Kenneth L., Alejandro Portes. 1980. "Immigrant Enclaves: An Analysis of the Labor Market Experiences of Cubans in Miami." American Journal of Sociology 86:295-319. Wilson, William Julius. 1987. The Truly Disadvantaged. The Inner City, the Underclass, and Public Policy. Chicago and London: University of Chicago Press. —. 1988. "Social Policy and Minority Groups. What Might Have Been and What Might We See in the Future?" Pp. 231-252 in Divided Opportunities. Minorities, Poverty, and Social Policy, edited by Gary D. Sandefur and Marta Tienda. New York/London: Plenum Press Wilson, William Julius 1996. When Work Disappears. The World of the New Urban Poor. New York: Alfred A. Knopf. Wilson, William Julius, Robert Aponte, and Kathryn Neckerrnan. 1987. "Joblessness versus Welfare Effects. A Further Reexamination." Pp. 93-106 in The Truly Disadvantaged. The Inner City, the Underclass, and Public Policy, edited by William Julius Wilson. Chicago and London: University of Chicago Press. Winant, Howard. 2000. "Race and Race Theory." Annual Review of Sociology 26:169-185. —. 2001. The World is a Ghetto. Race and Democracy Since World War II. New York: Basic Books. Wright, Lawrence. 1990. "One Drop of Blood." in The New Yorker, Annals of Politics. 281 Wusten, Herman van der. 2001. " Verzuiling and Post- Verzuiling Politics in the Netherlands: Coping with Changing Diversity." Pp. 307-319 in Ethnic Minority and Interethnic Relations in Context. A Dutch-Hungarian Comparison, edited by Karen Phalet and Antal Orkény. Ashgate: Aldershot. Zhou, Yu. 2002. "New York: Caught under the Fashion Runway." Pp. 113-133 in Unravelling the Rag Trade: Immigrant Entrepreneurship in Seven World Cities, edited by Jan Rath. Oxford/New York: Berg. 282 MICHIGAN STATE UNIVERSII‘Y LIBRARIES III: ‘ “ I III ‘ ”I III III: III IIII III III: I I III 3 293 02845 8069 I I 1