TY - JOUR
AU - Ugreninov,, Elisabeth
AB - Abstract Birth order causes social inequality between siblings. In Western countries, earlier-born perform better than later-born. In non-Western countries, however, earlier-born generally perform worse than later-born. We use administrative data to compare birth order effects between the native population and Norwegian-born children of immigrants (450,864 individuals nested within 202,191 families). The outcome is based on lower secondary school points—a combined measurement of all marks at the end of compulsory education. We found negative effects of birth order on school points for Norwegian natives. This finding is in line with previous studies from Western countries using measures of educational achievement. Birth order effects amongst Norwegian-born children of immigrant parents vary according to maternal country of origin, with some displaying similar effects as natives and others displaying no birth order effects. These findings are surprising in light of the main theories on birth order, which are universalistic and do not predict group-differences. We argue that the universalistic theories need to be supplemented to account for our findings, and discuss the potential importance of cultural variation in family-related behaviours such as strategic parenting, kindergarten attendance, and peer-influence in high-achieving immigrant groups. Introduction Birth order effects have received widespread research attention over the last 15 years. Most studies document that earlier-born perform better than later-born in an array of socioeconomic outcomes in Western countries (Black, Devereux and Salvanes, 2005; Kristensen and Bjerkedal, 2007, 2010; Barclay, 2014; Härkönen, 2014; Barclay, Hällsten and Myrskylä, 2016; Black, Grönqvist and Öckert, 2016). This research has shown that birth order effects cause socioeconomic differences within families, an important part of the overall picture of social inequality (Conley, 2005). There are two main theories that attempt to explain birth order effects: the resource dilution theory (Blake, 1981, 1986, 1989) and the confluence theory (Zajonc and Markus, 1975; Zajonc, 1976, 1983). These theories explain birth order effects as arising from parental resource allocation between siblings and between-sibling differences in the intellectual climate of the family, respectively. Both theories explain birth order effects in a universalistic manner; implying that one should expect to find negative birth order differences in nuclear families in any context. Research based on data from some non-Western countries have provided evidence for negative birth order effects (Abafita and Kim, 2015; Calimeris and Peters, 2017; Fors and Lindskog, 2017). However, many studies find positive birth order effects (Ejrnæs and Pörtner, 2004; Tenikue and Verheyden, 2010; De Haan, Plug and Rosero, 2014; Kumar, 2016) as well as non-linear or no birth order effects (Dayioǧlu, Kirdar and Tansel, 2009; Dammert, 2010; Seid and Gurmu, 2015) on socioeconomic outcomes. These studies encompass outcomes such as pre-school cognition, school enrolment, school attendance, and years of completed education. These findings contrast those of the studies based on Western data and therefore pose a challenge to the universalistic birth order theories. In some cases the absence of negative birth order effects are likely to be linked to contextual factors such as poverty and child labour that are less relevant in developed countries. On the other hand, the disparity may be related to differences in family practices across different countries. Importantly, such differences are likely to be exhibited by immigrants in their new host countries. In this article, we examine birth order effects in immigrant families in Norway in an attempt to unpack parts of this puzzle. This is important for a number of reasons. First, by studying families that have immigrated we contribute to a better understanding of birth order effects. In particular, we want to investigate whether the universalistic theories hold in this context and whether there is evidence of any cultural variation. Second, to the best of our knowledge, there have been no studies on birth order effects in immigrant families. We do not know if within-family differences in the form of birth order effects actually exist in these families, and within-family differences are an important source of social inequality. Third, most Western populations are characterized by sub-replacement fertility (Goldstein, Lutz and Testa, 2004; Lesthaeghe, 2010). Population growth is therefore being upheld by immigrants (Population and Population Change Statistics—Statistics Explained, 2017), who typically arrive at a young age and have higher fertility rates than the native populations (Sobotka, 2008). This difference in fertility rate implies that immigrants and their descendants comprise an increasing proportion of Western populations. A consequence is that immigrant families have more children with higher birth orders, and birth order effects could have a substantial negative impact on later-born children among immigrants. This in turn could influence the average socioeconomic outcomes of minority groups on an aggregated level. Previous Research and Theory Birth order has an impact on a range of outcomes such as income (Behrman and Taubman, 1986; Lindahl, 2008; Bertoni and Brunello, 2016), intelligence (Kristensen and Bjerkedal, 2007; Boomsma et al., 2008; Black, Devereux and Salvanes, 2011; Barclay, 2014; Sundet, 2014), and educational achievement (Black, Devereux and Salvanes, 2005; Booth and Kee, 2009; de Haan, 2010; Kristensen and Bjerkedal, 2010; Silles, 2010; Härkönen, 2014; Barclay, 2015). In the last 15 years, studies based on rigorous analytical approaches and high quality data have confirmed the existence of birth order effects beyond a reasonable doubt. This contrasts with the period before the 2010s when debates raged over the claim that birth order effects result from statistical designs that cannot differentiate causally between birth order and family size/Flynn-effects (Rodgers et al., 2000; Rodgers, 2001, 2014; Wichman, Rodgers and Maccallum, 2007). Birth order effects are real, but there are still many unanswered questions. The vast majority of birth order studies have used data from Western countries, typically utilizing samples that do not take variation according to immigrant status into account. As Western countries are arguably culturally homogenous, it is perhaps not surprising that theoretical explanations point to general mechanisms and that deriving expectations about group-differences is difficult based on most existing theories. There are several theories about birth order, but two in particular receive the most scholarly interest: Resource dilution theory and confluence theory. The starting point of the resource dilution theory is that parents have a finite amount of resources (Blake, 1981, 1986, 1989). The theory refers to three main types of resources: (i) physical environment—cultural objects, a home, etc.; (ii) opportunities—to engage with the outside world, have a walk, etc.; and (iii) treatments—personal attention, teaching, etc. (Blake, 1981). The latter is probably the most important for educational outcomes (Conley, 2005). The amount of available resources varies with birth order, as more children means that parents have more mouths to feed and more brains to train. Relative to their later-born siblings, earlier-born children will have had more of their parents’ resources exclusively available and their parents will have spent more quality time with them. This affects their outcomes in later stages of life, where earlier-born children display higher levels of intelligence, educational attainment, and income. Some studies have indicated that earlier-born children receive more economic (de Haan, 2010) and cultural (Price, 2008) resources from their parents. Still, the evidence pointing towards resource dilution as the prime driver of birth order effects is far from conclusive. Other studies find that intra-familiar differences in the allocation of resources do not necessarily explain birth other effects fully, and that other factors need to be taken into account (Monfardini and See, 2012; Workman, 2017). The confluence theory is built upon the same rationale as the resource dilution theory, with a slight twist. According to this theory, it is not the resources that matter, but the psychological climate in the family (Zajonc, 1983). Each additional child born implies a lowering of the overall intellectual environment in the family, affecting all siblings (Zajonc and Markus, 1975). As in the resource dilution theory, earlier-born children will initially have spent more time in a higher intellectual climate. With the arrival of a second-born sibling the intellectual climate in the family is lowered for the first-born. After the birth of the second sibling, however, the intellectual climate is better for the second-born compared to the first-born. This is because the intellectual climate for the second-born is the average of both parents and the first-born, whereas the intellectual climate for the first-born is the average of both parents and the younger second-born sibling. In addition, a ‘teaching-effect’ is also proposed (Zajonc and Sulloway, 2007), which implies that older siblings learn more from teaching their younger siblings, than younger siblings benefit from being taught. This ‘tutoring effect’ outweighs that the second-borns experience a more stimulating intellectual climate as the children become older. Therefore, first-born excel socioeconomically later in life compared to their younger siblings. Spacing between siblings also matters, and the birth order effects can be nullified or even reversed if the age gaps are big enough (Zajonc, 1976). This nullification is theoretically predicted to occur if the age gap between siblings is 11 ± 2 years. Another set of theories explain birth order effects as a result of biological differences arising in the womb (Gualtieri and Hicks, 1985; Almond and Currie, 2011). However, these theories have been dealt a blow in recent years (Kristensen and Bjerkedal, 2007, 2010; Barclay, 2015; Black, Devereux and Salvanes, 2015; Brenøe and Molitor, 2017). Other theories include a more general role- or personality-theory (e.g. Sulloway, 1996; Conley, 2005), as well as theories emphasising strategic parenting that lead to differential treatment of children according to birth order (Hotz and Pantano, 2015; Pavan, 2016). Mothers may also display behavioural differences at the pre-natal stage. For example, some studies indicate that mothers are more likely to exhibit potentially detrimental behaviours such as smoking during the pregnancy and breastfeeding less at higher birth orders (Lundberg et al., 2009; Buckles and Kolka, 2014). Finally, a cultural origin of birth order effects in (parts of) the West has been proposed, as such effects could reflect behaviours associated with a cultural remnant of primogeniture (Steelman et al., 2002; Kristensen and Bjerkedal, 2010; Barclay, Hällsten and Myrskylä, 2016). Earlier-born, and particularly first-born (sons), would benefit most and be prioritized in a form of strategic parenting. The main theories for the observed negative birth order patterns are not very helpful for explaining variation across groups. The resource dilution and confluence theories in particular are universal in their forms, and one should expect to find the same kind of patterns in immigrant families as in native population families. Still, the last set of theories postulate mechanisms that might vary depending on the social context. Other strands of research, however, point to the importance of a broader set of immigrant-related factors and suggest some reasons why we should expect differences in birth order effects between the immigrant families we study and the native population. Birth Order and Immigration The literature on birth order effects and education on natives in non-Western countries is growing. In recent years, important work has considered birth order effects in a range of countries. In a study looking at 12 countries in sub-Saharan Africa (Tenikue and Verheyden, 2010) birth order effects on educational attainment were documented in most of the countries, but they differ from those typically found for Western countries. In poor families there is a positive birth order effect on educational attainment. In rich families there is a negative effect. The authors link this to child labour: the oldest siblings in poor families have to work, which has a detrimental effect on their participation in school and less parental investment in their education. On the other hand, this does not apply to rich families. Additionally, Abafita and Kim (2015) provide evidence for negative birth order effects on education in Ethiopia, whilst Seid and Gurmu (2015) found no such effects in the same country. In Ecuador (De Haan, Plug and Rosero, 2014) there is a similar pattern as in sub-Saharan Africa (Tenikue and Verheyden, 2010) with both positive and negative birth order effects on educational attainment for poor and rich households, respectively. A study looking at Nicaragua and Guatemala (Dammert, 2010) found no effects on education, but negative birth order effects on child labour. In the Philippines there is evidence of positive birth order effects on both the amount of schooling and educational attainment (Ejrnæs and Pörtner, 2004). As for Indonesia (Calimeris and Peters, 2017) there are indications of negative birth order effects on cognitive ability, but the trend is not uniform. All the aforementioned studies use data for countries that are not included in our immigrant sample. However, three of the countries of origin included in our study are covered by studies on birth order and education. In Turkey, birth order effects on school enrolment are non-linear with middle-born children faring worse than their elder siblings, and the effect is more marked in poor families (Dayioǧlu, Kirdar and Tansel, 2009). In India, Fors and Lindskog (2017) found negative birth order effects on grades and parental investment into school quality. However, Makino (2018) and Kumar (2016) found positive birth order effects on school test scores and educational attainment. Anh et al. (1998) mention that they have performed a sensitivity analysis on birth order and education in Vietnam, but the results are not shown. According to the authors there were no birth order effects on school attendance and educational attainment. While a few of these studies provide evidence for negative birth order patterns that are closer to the birth order effect found in native Western populations, the majority document a non-existing or positive birth order effect on educational outcomes. This is not in line with the expectation based on the resource dilution and confluence theories. This finding points to the importance of considering institutional settings and a broader cultural context when examining birth order effects. Universal approaches need to consider possible institutional or cultural variations. We note that the aforementioned studies on Turkey, India, and Vietnam found non-linear, positive birth effects, or no birth order effects. If there is cultural variation at play in birth order effects in immigration families in Norway, these three studies allows us to theoretically compare if origin effects and Norwegian-born to Turkish, Indian, and Vietnamese parents. Institutional Settings and Poverty If the positive or non-existing birth order effects in non-Western countries are driven primarily by families who live in poverty, we do not expect to find the same in immigrant families in Norway. First, while Norwegian immigrants typically earn less than the native population and are more likely to enter lower social strata, the Norwegian welfare system is generous and immigrants are generally much better off in Norway compared to their country of origin. In addition, having a lower socioeconomic status in Norway is not likely to have the same relevance for the order and magnitude of birth order effects as it would have in one’s country of origin. In line with this, previous studies indicate that there are no differences in birth order effects on education according to parental socioeconomic status in Western countries (Black, Devereux and Salvanes, 2005; Härkönen, 2014). Also, the small subset of studies where differences according to parental socioeconomic status are found do not point in any particular direction. For example, some research provides evidence for larger birth order effects in families with low parental socioeconomic status, while others document the opposite (Bonesrønning and Massih, 2011; Barclay, Hällsten and Myrskylä, 2016; Grätz, 2018). Second, several studies from non-Western countries also consider different outcomes, such as child labour, that are closely related to poverty as well as institutional arrangements that are far less relevant in Norway and most Western countries. When child labour is the outcome variable, researchers typically document a negative birth order effect, in contrast to the positive association between birth order and educational outcomes in the same samples. The two are likely to be linked, since earlier-born children are required to work to help their families and are unable to attend school (Tenikue and Verheyden, 2010; De Haan, Plug and Rosero, 2014). However, child labour is prohibited by law in Norway, and the different economic circumstances make it less relevant in any case. To sum up, if institutional settings related to poverty are the main explanation for the different findings in non-Western and Western countries, we do not expect to see any variation according to parental origin. More specifically, we do expect immigrant families to display birth order patterns similar to the native population, as resource constraints are no longer as pressing upon arrival in Norway, and practices such as child labour are both redundant and prohibited. Cultural Variation in Family-Related Behaviours The negative birth order effects documented in Western countries might have a cultural origin, more specifically related to the practice of primogeniture. The oldest (son) would typically inherit his parent’s main property. A cultural remnant of this historical tradition might be relevant for explaining the general preference for first-borns, although there is little evidence indicating that first-born boys are advantaged compared to girls with the same parity (Kristensen and Bjerkedal, 2010; Barclay, Hällsten and Myrskylä, 2016). Along the same vein, some researchers have proposed explanations of positive birth order effects in non-Western countries that are linked to various cultural practices. For example, De Haan, Plug and Rosero (2014) pointed to the increase of breastfeeding frequency and quality time with birth order. This is the opposite of what is typically found in studies from Western countries (Buckles and Kolka, 2014). Such explanations might be relevant for minority families in Norway as well, since immigrants bring cultural practices, preferences, and beliefs from their counties of origin (Almond, Edlund and Milligan, 2013; Wimmer and Soehl, 2014; Kavli, 2015). Previous studies indicate that some family-related behaviours and preferences of immigrants change little in response to the different institutional incentives present in Norway. This is true for a broad set of family behaviours, for example, related to the distribution of labour inside and outside the household as well as systems of kinship and residence (Heath, Rothon and Kilpi, 2008). Fertility behaviour is another important example. While there is some evidence that childbirth rates decline somewhat with length of residence (Lappegård, 2006), immigrants’ preferences for their children’s sexes do not change much with length of residence in Norway (Lillehagen and Lyngstad, 2018). The same pattern has also been documented in other countries (Almond, Edlund and Milligan, 2013). Following this line of thought, a cultural persistence of a broad set of family behaviours might influence birth order effects. If there is variation in family-related behaviours between parents with immigrant and native origin, we expect to find group-differences. The Norwegian Context: Immigration to a Welfare State The history of immigration in Norway is similar to several other Western European countries in many respects (Brochmann and Kjeldstadli, 2008). Modern-day immigration to Norway from non-European countries began around 1970 primarily as labour immigration. Moroccan, Pakistani, and Turkish men were the largest groups at the time. After the so-called immigration stop in 1975, this large-scale, low-skilled labour immigration came to a halt. High-skilled labour immigration continued to occur, and other types of immigration became more important. Refugees and asylum seekers from conflict areas have immigrated from the 1970s to present day, with the largest groups originating in Chile, Vietnam, Sri Lanka, Iran, the Balkans, Somalia, Iraq, and recently, Syria. Family reunification has been important throughout. As a result of these processes, present day Norway is an ethnically diverse society (Brochmann and Kjeldstadli, 2008). As in many European countries, the number of immigrants from non-European countries and native-born to immigrant parents from non-European countries is rising, reaching 10 per cent in Norway in 2017 (Statistics Norway, 2018). The emergence of non-European immigration to Norway coincided with the consolidation of what has been known as the Nordic model. Esping-Andersen (1990) described the Nordic countries as having a social democratic welfare regime. Nordic countries are characterized by low levels of economic inequality, high gender equality, and high social mobility (Kautto, 2010). The Norwegian welfare state is universal and provides all citizens with equal social rights, such as near-free health care services, rights to welfare benefits, and free access to education. Legal immigrants are granted the same rights as everyone else, and the Norwegian-born children to immigrant parents included in this study have grown up in the same institutional setting as native children from the same cohorts. Some of these rights and duties are directly related to our outcome variable. Educational rights are the same for everyone and are heavily subsidized. All children have a right to kindergarten (ages 1–5), a statutory duty to attend primary and lower secondary school (ages 6–16), a statutory right to high school (ages 16–19), and all adults have equal access to higher education. Another important point is that immigrants in Norway are, comparatively speaking, well-integrated in the education system and labour market. Many groups have levels of education on par with natives and relatively high levels of occupational attainment (Brekke and Mastekaasa, 2008; Hermansen, 2013). Data and Methods This study used administrative data encompassing the total population of Norway. We linked information from the Central Population Registry to registries containing educational information. Each individual had an identification number, which made it possible to link parents with their children. To be included in our sample, the identities of both the father and the mother had to be known. We excluded children of plural births, as well as those with only half-siblings, as we are interested in a nuclear family setting. Importantly, only individuals registered with a non-missing value for lower secondary school points from 2001 to 2014 were eligible for the analytical sample. We performed several sensitivity analyses which clearly indicated that none of these decisions affected our main findings. Our final sample covers the 1985–1998 cohorts, which encompassed the period in which many Norwegian-born children of immigrant parents who arrived in the 1970s and 1980s were coming of age. Furthermore, to meet the data requirements for the family fixed effects model, only individuals from families in which at least two siblings meet the sample requirements were included in our analysis. Variables The outcome variable is lower secondary school points. At the end of compulsory education, pupils are given a combined measurement for all marks, which determines which upper secondary schools he or she will be able to attend. The school points measure is based on a combination of both final assessment marks and examination marks. The grading system was changed in 2007, however. We accounted for this, as well as potential grade inflation, by standardizing the lower secondary school point variable by year. Birth order was included as a set of dummies with values from one to four, with first-borns as the reference category. We excluded information on individuals with birth orders higher than four, as the low number of observations in these categories for some groups entails privacy issues. Families with more than four children are still included. Fourth-borns in this study are therefore not necessarily last-borns, as some of them may have younger siblings which have been excluded. Due to the sample restrictions described above there are more second-borns than first-borns. This is because first-borns more often fall outside the observation period of the outcome variable. This has a similar impact on the number of observations across groups with different origins. We performed a sensitivity analysis in which only families with valid values for first-borns were included (see Supplementary Table SA1 and Figure SA1). These results are very similar to those from our preferred model. We divide into groups based on maternal country of origin. The native group encompassed individuals born in Norway to Norwegian-born parents. The immigrant groups consisted of children with two foreign-born parents. We use the mother’s country of birth to measure the country of origin. We exclude children from families with mixed origin, e.g. with one immigrant parent and one native parent or parent from a different country. In addition to children born in Norway, we also include children who immigrated with their parents when they were 5 years or younger in order to obtain a slightly larger sample and more precise estimates. This addition is reasonable as it has been shown that childhood immigrants are similar to the native population in terms of later socioeconomic outcomes when arriving before the age of six (Steinkellner, 2013; Hermansen, 2017). In an additional analysis, we split the sample based on age at arrival (see Supplementary Figure SA1) and found that the magnitude of the estimates did not depend on whether this group was included or not. Previous power-tests done by Bjerkedal et al. (2007), as well as our own indicated that one needs around 700–900 sibling pairs to be able to detect birth order effects of the magnitude typically found in Norway. After making the necessary sample restrictions, we divide the sample into immigrant groups where there are enough observations left to be in that range. This gives us 11 groups: Natives (Norway), Chile, India, Iran, Iraq, Morocco, Pakistan, Somalia, Sri Lanka, Turkey, and Vietnam. We also include those of European origin as a regional group for comparison as these are quite culturally homogenous. Sex was included as a dummy-variable to adjust the birth order coefficients, as there is a difference in school points between girls and boys. We perform the analysis both with and without for the inclusion of individual year of birth dummies to compare the two results. We do this to see if length of parental residence influences the birth order effects in our immigrant samples, as previous research indicates that this could have a positive effect on children’s educational attainment (Smith, Helgertz and Scott, 2016). Furthermore, when included in the fixed effects model controls for year of birth serves as an adjustment for other secular trends as well, such as increased maternal and paternal age by birth order. It thereby adjusts for possible changes at the parental level such as (increased) parental income level, general maturation and (decreasing) health as well. This also allows us to investigate whether increasing resources at the parental level matter more in immigrant families than in native families. This variable is interacted with the origin variable in order to allow the effect of such factors to vary between groups. Descriptive Statistics Table 1 provides summary statistics with means and standard deviations for each variable. After making the restriction described above, we are left with 450,864 individuals within 202,191 families. Table 1. Descriptive statistics . 1 . 2 . 3 . 4 . 5 . 6 . 7 . 8 . 9 . 10 . 11 . 12 . Mothers country of origin Chile Europe India Iran Iraq Morocco Native Pakistan Somalia Sri Lanka Turkey Vietnam School point (mean) 36.64 41.08 43.50 40.27 35.86 37.62 42.16 38.47 36.58 42.96 35.02 42.07 School point (SD) 8.57 9.22 8.10 8.87 8.69 8.25 8.75 8.59 7.58 7.87 8.58 8.19 1st born 232 2,276 347 273 203 237 166,202 1,147 230 594 578 845 1st born (per cent) 34 36 39 35 26 30 39 26 29 40 34 31 2nd born 280 2,601 382 336 271 260 183,513 1,404 273 624 656 1,004 2nd born (per cent) 41 41 43 43 35 33 43 32 34 42 39 37 3rd born 141 1,105 123 124 181 184 64,906 1,105 190 217 325 569 3rd born (per cent) 21 17 14 16 23 23 15 25 24 15 19 21 4th born 34 396 33 50 121 107 13,690 768 114 36 125 277 4th born (per cent) 5 6 4 6 16 14 3 17 14 2 7 10 Average family size 2.86 2.97 2.69 2.84 4.36 4.04 2.71 4.09 4.96 2.84 3.26 3.26 Average spacing 3.75 2.97 2.97 3.70 2.46 2.96 3.20 2.61 2.07 3.07 3.42 2.83 Female (per cent) 47 50 50 49 50 49 49 50 49 51 50 49 Birth year (mean) 1990.78 1992.22 1992.15 1991.72 1993.68 1992.05 1991.58 1991.46 1993.76 1993.10 1991.96 1991.84 Birth year (SD) 3.67 3.48 3.58 3.79 3.35 3.73 3.68 3.85 3.39 3.48 3.79 3.63 Mothers age at birth (mean) 28.00 27.88 28.41 27.21 25.58 26.97 28.05 26.71 27.04 28.91 25.13 27.52 Mothers age at birth (SD) 4.69 4.86 4.78 5.19 4.65 4.95 4.58 4.60 4.28 4.52 4.71 4.88 Mothers length of residence (mean) 3.57 6.74 5.47 0.96 −0.94 5.96 27.00 5.31 1.38 3.59 5.38 4.83 Mothers length of residence (SD) 5.30 11.00 5.90 3.88 3.20 5.04 3.93 5.38 3.57 3.99 5.45 5.22 Families 303 2,818 401 379 346 321 190,755 1,760 335 646 727 1,148 Observations 687 6,378 885 783 776 788 428,311 4,424 807 1,471 1,684 2,695 . 1 . 2 . 3 . 4 . 5 . 6 . 7 . 8 . 9 . 10 . 11 . 12 . Mothers country of origin Chile Europe India Iran Iraq Morocco Native Pakistan Somalia Sri Lanka Turkey Vietnam School point (mean) 36.64 41.08 43.50 40.27 35.86 37.62 42.16 38.47 36.58 42.96 35.02 42.07 School point (SD) 8.57 9.22 8.10 8.87 8.69 8.25 8.75 8.59 7.58 7.87 8.58 8.19 1st born 232 2,276 347 273 203 237 166,202 1,147 230 594 578 845 1st born (per cent) 34 36 39 35 26 30 39 26 29 40 34 31 2nd born 280 2,601 382 336 271 260 183,513 1,404 273 624 656 1,004 2nd born (per cent) 41 41 43 43 35 33 43 32 34 42 39 37 3rd born 141 1,105 123 124 181 184 64,906 1,105 190 217 325 569 3rd born (per cent) 21 17 14 16 23 23 15 25 24 15 19 21 4th born 34 396 33 50 121 107 13,690 768 114 36 125 277 4th born (per cent) 5 6 4 6 16 14 3 17 14 2 7 10 Average family size 2.86 2.97 2.69 2.84 4.36 4.04 2.71 4.09 4.96 2.84 3.26 3.26 Average spacing 3.75 2.97 2.97 3.70 2.46 2.96 3.20 2.61 2.07 3.07 3.42 2.83 Female (per cent) 47 50 50 49 50 49 49 50 49 51 50 49 Birth year (mean) 1990.78 1992.22 1992.15 1991.72 1993.68 1992.05 1991.58 1991.46 1993.76 1993.10 1991.96 1991.84 Birth year (SD) 3.67 3.48 3.58 3.79 3.35 3.73 3.68 3.85 3.39 3.48 3.79 3.63 Mothers age at birth (mean) 28.00 27.88 28.41 27.21 25.58 26.97 28.05 26.71 27.04 28.91 25.13 27.52 Mothers age at birth (SD) 4.69 4.86 4.78 5.19 4.65 4.95 4.58 4.60 4.28 4.52 4.71 4.88 Mothers length of residence (mean) 3.57 6.74 5.47 0.96 −0.94 5.96 27.00 5.31 1.38 3.59 5.38 4.83 Mothers length of residence (SD) 5.30 11.00 5.90 3.88 3.20 5.04 3.93 5.38 3.57 3.99 5.45 5.22 Families 303 2,818 401 379 346 321 190,755 1,760 335 646 727 1,148 Observations 687 6,378 885 783 776 788 428,311 4,424 807 1,471 1,684 2,695 Note: Family size calculated after exclusion of one-child families, but before other sample restrictions. Source: Authors own calculations based on Norwegian registry data. Open in new tab Table 1. Descriptive statistics . 1 . 2 . 3 . 4 . 5 . 6 . 7 . 8 . 9 . 10 . 11 . 12 . Mothers country of origin Chile Europe India Iran Iraq Morocco Native Pakistan Somalia Sri Lanka Turkey Vietnam School point (mean) 36.64 41.08 43.50 40.27 35.86 37.62 42.16 38.47 36.58 42.96 35.02 42.07 School point (SD) 8.57 9.22 8.10 8.87 8.69 8.25 8.75 8.59 7.58 7.87 8.58 8.19 1st born 232 2,276 347 273 203 237 166,202 1,147 230 594 578 845 1st born (per cent) 34 36 39 35 26 30 39 26 29 40 34 31 2nd born 280 2,601 382 336 271 260 183,513 1,404 273 624 656 1,004 2nd born (per cent) 41 41 43 43 35 33 43 32 34 42 39 37 3rd born 141 1,105 123 124 181 184 64,906 1,105 190 217 325 569 3rd born (per cent) 21 17 14 16 23 23 15 25 24 15 19 21 4th born 34 396 33 50 121 107 13,690 768 114 36 125 277 4th born (per cent) 5 6 4 6 16 14 3 17 14 2 7 10 Average family size 2.86 2.97 2.69 2.84 4.36 4.04 2.71 4.09 4.96 2.84 3.26 3.26 Average spacing 3.75 2.97 2.97 3.70 2.46 2.96 3.20 2.61 2.07 3.07 3.42 2.83 Female (per cent) 47 50 50 49 50 49 49 50 49 51 50 49 Birth year (mean) 1990.78 1992.22 1992.15 1991.72 1993.68 1992.05 1991.58 1991.46 1993.76 1993.10 1991.96 1991.84 Birth year (SD) 3.67 3.48 3.58 3.79 3.35 3.73 3.68 3.85 3.39 3.48 3.79 3.63 Mothers age at birth (mean) 28.00 27.88 28.41 27.21 25.58 26.97 28.05 26.71 27.04 28.91 25.13 27.52 Mothers age at birth (SD) 4.69 4.86 4.78 5.19 4.65 4.95 4.58 4.60 4.28 4.52 4.71 4.88 Mothers length of residence (mean) 3.57 6.74 5.47 0.96 −0.94 5.96 27.00 5.31 1.38 3.59 5.38 4.83 Mothers length of residence (SD) 5.30 11.00 5.90 3.88 3.20 5.04 3.93 5.38 3.57 3.99 5.45 5.22 Families 303 2,818 401 379 346 321 190,755 1,760 335 646 727 1,148 Observations 687 6,378 885 783 776 788 428,311 4,424 807 1,471 1,684 2,695 . 1 . 2 . 3 . 4 . 5 . 6 . 7 . 8 . 9 . 10 . 11 . 12 . Mothers country of origin Chile Europe India Iran Iraq Morocco Native Pakistan Somalia Sri Lanka Turkey Vietnam School point (mean) 36.64 41.08 43.50 40.27 35.86 37.62 42.16 38.47 36.58 42.96 35.02 42.07 School point (SD) 8.57 9.22 8.10 8.87 8.69 8.25 8.75 8.59 7.58 7.87 8.58 8.19 1st born 232 2,276 347 273 203 237 166,202 1,147 230 594 578 845 1st born (per cent) 34 36 39 35 26 30 39 26 29 40 34 31 2nd born 280 2,601 382 336 271 260 183,513 1,404 273 624 656 1,004 2nd born (per cent) 41 41 43 43 35 33 43 32 34 42 39 37 3rd born 141 1,105 123 124 181 184 64,906 1,105 190 217 325 569 3rd born (per cent) 21 17 14 16 23 23 15 25 24 15 19 21 4th born 34 396 33 50 121 107 13,690 768 114 36 125 277 4th born (per cent) 5 6 4 6 16 14 3 17 14 2 7 10 Average family size 2.86 2.97 2.69 2.84 4.36 4.04 2.71 4.09 4.96 2.84 3.26 3.26 Average spacing 3.75 2.97 2.97 3.70 2.46 2.96 3.20 2.61 2.07 3.07 3.42 2.83 Female (per cent) 47 50 50 49 50 49 49 50 49 51 50 49 Birth year (mean) 1990.78 1992.22 1992.15 1991.72 1993.68 1992.05 1991.58 1991.46 1993.76 1993.10 1991.96 1991.84 Birth year (SD) 3.67 3.48 3.58 3.79 3.35 3.73 3.68 3.85 3.39 3.48 3.79 3.63 Mothers age at birth (mean) 28.00 27.88 28.41 27.21 25.58 26.97 28.05 26.71 27.04 28.91 25.13 27.52 Mothers age at birth (SD) 4.69 4.86 4.78 5.19 4.65 4.95 4.58 4.60 4.28 4.52 4.71 4.88 Mothers length of residence (mean) 3.57 6.74 5.47 0.96 −0.94 5.96 27.00 5.31 1.38 3.59 5.38 4.83 Mothers length of residence (SD) 5.30 11.00 5.90 3.88 3.20 5.04 3.93 5.38 3.57 3.99 5.45 5.22 Families 303 2,818 401 379 346 321 190,755 1,760 335 646 727 1,148 Observations 687 6,378 885 783 776 788 428,311 4,424 807 1,471 1,684 2,695 Note: Family size calculated after exclusion of one-child families, but before other sample restrictions. Source: Authors own calculations based on Norwegian registry data. Open in new tab The school point average varies somewhat between the groups. It is also worth noting that Norwegian-born children of immigrant parents from Europe, India, Sri Lanka, and Vietnam perform on par with or better than natives. Higher birth orders are more common in immigrant families from Iraq, Morocco, Pakistan, and Somalia. Method Our analyses were based on what is considered the most rigorous strategy in the field of birth order: linear regression models with fixed-effects clustered at the family level, i.e. sibling fixed-effect models. Clustering at the family level removes unobserved heterogeneity between families and therefore corrects for the possible confounding effects of family size. The estimated equation can be specified as: Yi= α + αi+ βXi+ γWi+ βFamXiFam+ γFamWiFam+ εi. (1) Observed (βFam XiFam) and unobserved (γFam WiFam ⁠) traits or behaviours that affect all siblings in the same family equally were factored out. We estimate two main models. Model 1 includes dummies for birth orders one to four, country of origin, and a sex. In Model 2 we also include year of birth of each child as a set of dummy variables, which are interacted with the origin dummies to allow for heterogeneity. In both Model 1 and Model 2, the estimates for the birth order effects for each origin group are derived from a pooled model were birth order is interacted with origin. The birth order coefficients for each origin group are plotted in Figures 1 and 2, obtained by varying the reference category of the origin variable and otherwise using the same model specification. The number of groups (families) was large, so we used cluster-robust standard errors, based on the assumption that the groups were independent of each other but that the individuals in each group were not. Figure 1. Open in new tabDownload slide Birth order and z-scores for lower secondary school points by maternal country/region of origin. Without control for cohort Note: The round points mark birth order effects in the native population. Fixed effect regression models with clustering at the family level and controls for sex. Figure 1. Open in new tabDownload slide Birth order and z-scores for lower secondary school points by maternal country/region of origin. Without control for cohort Note: The round points mark birth order effects in the native population. Fixed effect regression models with clustering at the family level and controls for sex. Figure 2. Open in new tabDownload slide Birth order and z-scores for lower secondary school points by parental region/country of origin. With control for cohort Note: The round points mark birth order effects in the native population. Fixed effect regression models with clustering at the family level and controls for cohort and sex. Figure 2. Open in new tabDownload slide Birth order and z-scores for lower secondary school points by parental region/country of origin. With control for cohort Note: The round points mark birth order effects in the native population. Fixed effect regression models with clustering at the family level and controls for cohort and sex. Note that we utilize statistical significance tests and present confidence intervals even though we have information on the complete population due to our use of registry-based administrative data. We base this on statistical arguments which are well presented in Aaberge and Laake (1984) and Hoem (2008). Additionally, this is standard practice in social science research based on administrative data. When using stochastic models, the population can be seen as one of many possible realizations of the underlying mechanisms that are studied. In other words, individual life stories should be seen as realizations of stochastic processes subject to random variation, and that such variation should be taken into account even when the set of observations contain all members of a population (Hoem 2008: p. 439). Results Figure 1 displays the main results for all eleven origin groups based on the within-family regression (full regression tables are available in Supplementary Table SA2) without individual controls for year of birth. The lower secondary school point average of first-borns serve as a reference, marked by the black solid horizontal line. The plotted coefficients for the native population are displayed in red/grey with a round point and are presented in each panel for comparison. In the native population, there is a negative relationship between birth order and lower secondary school points. Many groups of immigrant families (India, Pakistan, Turkey, and Vietnam) display statistically significantly weaker negative effects of birth order on lower secondary school points compared to the native families. Meanwhile, other groups display birth order effects similar to the native population (Chile, Europe, Iran, Iraq, Morocco, and Somalia). In Figure 2 individual controls for year of birth for each child is included in the models and interacted with the origin variable to allow for heterogeneity between groups (for full regression tables, see Supplementary Table SA3). This enables us to adjust for variation which could be linked to time-varying factors such as (increasing) family income, (decreasing) parental health, as well as inter-sibling differences in parental length of residence, which is only relevant for immigrant families. Previous research utilizing a family fixed effects design indicates that parental time of residence is positively associated with the school success of immigrant children in Sweden, but that there is heterogeneity between groups of different origin (Smith et al., 2016). Therefore, this model allows us to isolate mechanisms that are specific to birth order to a greater extent. Amongst natives, the average difference between first-borns and second-borns is approximately 2 school points (22 per cent of a standard deviation), and 3.2 (36 per cent of a standard deviation) and 4 school points (46 per cent of a standard deviation) for third-borns and fourth-borns compared to first-borns, respectively. This negative relationship is consistent with findings from previous studies on birth order and education in Norway (e.g. Black, Devereux and Salvanes, 2005; de Haan, 2010; Kristensen and Bjerkedal, 2010), although different measures of schooling outcomes were used. Our results provide further evidence that differences in educational achievement by birth order are present in adolescence and in the formative years of lower secondary school for the native population. The difference between first-borns and fourth-borns is of a magnitude similar to that of the sex gap in our sample, where girls outperform boys by approximately 4 school points (48 per cent of a standard deviation) on average. The difference between birth orders one and four is, in other words both statistically and substantially significant. However, our main interest is birth order effects in immigrant families. Adjusting for year of birth provides us with a slightly different picture. The estimates for the native population changes somewhat, with the negative effects of having a higher birth order becoming slightly more marked. Immigrants originating in Chile, Europe, Iran, Iraq, Morocco, and Somalia still show negative birth order effects that are not statistically discernible from those of the native population (with high statistical uncertainty for some groups). Immigrant families of Turkish origin now display negative birth order effects that are similar to those of natives. Children in immigrant families of Pakistani origin still display significantly weaker negative birth order effects compared to natives, but the effects are now statistically different from zero. These changes are consistent with individual year of birth controls adjusting for parental time of residence and other cohort-related factors that affect the educational success of later-born siblings within the same family positively. Importantly, immigrant families originating in Sri Lanka and Vietnam display no birth order effects. Children from families of Indian origins display a positive (though not statistically significant) relationship between birth order and school points when controls for year of birth are included. Additionally, results from an alternative specification, where linear birth order was interacted with origin (with native families as the reference category), confirms that the overall trend in the birth order effects families originating in Pakistan, Sri Lanka, India, and Vietnam are significantly less negative than those found in native families (see Supplementary Table SA4). Discussion In previous research indicating birth order effects on educational achievement in Western countries, later-born children are more likely to have lower educational achievement compared to older siblings (Black, Devereux and Salvanes, 2005; Kristensen and Bjerkedal, 2007, 2010; Barclay, 2014, 2016; Härkönen, 2014). In this study, we examined birth order effects in immigrant families. Using high quality data, our results indicated a negative birth order effect for the native population in Norway, which corresponds well with previous research. Among Norwegian-born children of immigrant parents, the picture is more complicated. Many groups of immigrant families have birth order effects similar to the native population, but there are important exceptions. Notably, the results for India, Turkey, and Vietnam in our immigrant sample is quite similar to the findings in studies on natives in these countries (Anh et al., 1998; Dayioǧlu, Kirdar and Tansel, 2009; Kumar, 2016; Makino, 2018), suggesting that there is cultural variation in birth order effects according to cultural origin. While we are not aware of any previous studies covering the native population of Sri-Lanka, our findings for this country could also point in the same direction. Most of the findings for the non-Western immigrant groups are consistent with what we previously referred to as the universalistic assumptions of the resource and dilution theory, meant to cover all (nuclear) families. In our preferred model, displayed in Figure 2, five of the nine non-Western immigrant groups have birth order estimates which are nearly identical to the native population of Norway. Some of our findings, however, seemed to contradict the universalistic implications of the resource theory and the confluence theory, as these theories predict similar effects across all groups. Our findings therefore indicate that one or more of these assumptions are less general than these theories imply. We would like to address this theoretical inconsistency by theorizing further on why this may be the case. Our suggestions are relevant for many immigrant families, but less so for the native population, and may therefore explain the group-differences we documented. Resource Dilution Theory According to the resource dilution theory, parents possess a finite amount of resources that they distribute among their young. Each birth adds more mouths and brains to feed, and this lays the groundwork for differences between siblings. Later-born children receive fewer resources than earlier-born children, as there are less available (Blake, 1981, 1986, 1989). In many immigrant families, this assumption may not be valid, as the amount of resources available to the parents could increase rapidly during the first few years after immigration (Sandnes, 2017). A rapid increase of family resources after immigration could therefore lead to a birth order pattern that is the opposite from that found for native populations. The amount of economic resources and cultural resources linked to knowledge about the new country of residence would increase with the birth order in this case, thereby providing younger siblings with more family resources than earlier-born children. If this process is particularly strong for some immigrant groups, as opposed to for natives, this could potentially explain our findings for India, Sri Lanka, and Vietnam. We investigated whether this was the case in different ways. First, by using a within-family estimation strategy and controlling for individual year of birth dummies interacted with origin, thereby allowing for heterogeneity (Figure 2). In doing this we were able to adjust for at least some of the variations linked to increased economic and cultural resources over birth orders at the parental level. We also tested this assumption more directly by an analysis that uses total family income in the year of birth of each sibling as the outcome, and birth order as the main explanatory variable which is interacted with origin (see Supplementary Table SA5). While there were some differences in the income trajectory by birth order between families of different origins, these do not match up with the differences in birth order effects from our main results. This evidence indicates that differences in the income trajectories over time for families of different origins do not explain group-differences in birth order effects. Instead, the difference in birth order effects between children of immigrant parents from different countries suggest that the variation is linked to group-specific factors (e.g. cultural or country-specific) rather than explanations linked to immigrant status itself. Confluence Theory The confluence theory states that differences in the overall intellectual climate in the family across siblings of different ages cause socioeconomic differences according to birth order. Earlier-born children experience longer exposure to a higher intellectual climate and have a positive ‘tutoring effect’ of teaching younger siblings, and fare better in socioeconomic terms later in life as a result. Spacing between siblings is important, and a spacing gap of around 11 years will nullify birth order effects (Zajonc, 1976, 1983). Although birth order theories are silent on variations based on immigrant background, one possibility is that the well-documented differences is fertility rates between natives and immigrants of different origins (Lappegård, 2006) leads to differences in birth intervals between groups, which could in turn affect the magnitude and direction of birth order effects. We investigated this in two ways. First, Table 1 includes the average birth interval for each origin group. While there are some differences in birth intervals, there seems to be no systematic co-variation between birth intervals and the birth order effects displayed in Figure 2 for immigrant groups. Second, we carried out an additional analysis with the number of years since the last sibling was born as the main independent variable, otherwise using the same specification as in our main models with family fixed effects (but excluding first-borns). We interacted this spacing variable with origin. The results of this analysis (presented in Supplementary Table SA6) show that the effects of spacing does not differ across origin groups. Taken together, these two sources of evidence strongly suggest that differences in birth spacing between groups cannot explain our overall findings. Sex Preferences and Strategic Parenting Many non-Western countries may have different cultural practices than Western, such as sex preferences (e.g. De Haan, Plug and Rosero, 2014; Lillehagen and Lyngstad, 2018). Previous research from Norway has indicated that some immigrants typically display a preference for boys (Lillehagen and Lyngstad, 2018). If cultural remnants of primogeniture lead native parents to invest more in earlier-born children regardless of sex, whereas non-Western immigrants primarily invest in (older) boys, this could lead to weaker overall birth order effects on education in the latter group. We tested this by interacting birth order with sex for all groups. There is a minuscule positive interaction-effect of being later-born female amongst natives. Also, for later-born children of Sri Lankan and Turkish descent there is a positive interaction-effect of being a later-born female. It should be noted, however, that these groups are very small (see Supplementary Table SA7 for details). Due to the high number of tests, some of these might be false positives, but in any case the significant interactions provide no systematic evidence for boys being favoured. Differences in sex preferences are therefore not likely to explain our results. A more general form of cultural variation could be linked to strategic parenting. For Western populations, studies have suggested that parents are more lenient toward later-born children, as indicated by lower levels of breastfeeding, higher levels of smoking during the pregnancy, etc. (Hotz and Pantano, 2015). On the other hand, studies from non-Western countries find that mothers tend to breastfeed later-born children longer and push their earlier-born children into the child labour market more often than their later-born children (Tenikue and Verheyden, 2010; De Haan, Plug and Rosero, 2014; Seid and Gurmu, 2015). This could be an indication of a form of strategic parenting that may play out differently for immigrant parents compared to parents in the native population. Kindergarten Cultural family orientations might cause different adaption to educational institutions among earlier- and later-born children, which is highly relevant for policymakers. The Norwegian welfare state promotes kindergarten as one of the most important tools for the development of language skills and cultural integration for immigrants in particular and reduces social differences and income poverty in general. In Norway, 77 per cent of children aged 0–5 attend kindergarten (Statistics Norway, 2017). Immigrant mothers from non-European countries are more likely to be full-time homemakers and are more sceptical of sending their children to kindergarten (Nadim, 2014). Earlier-born children in immigrant families may attend kindergarten at a later age than their younger siblings for a number of reasons. First, the parents may be more receptive of kindergarten education after earlier-born children have attended. Second, if earlier-born children are already attending kindergarten, parents may decide to enrol their younger siblings for practical and time-saving reasons, thus enrolling later-born children at a younger age than their older siblings. Third, children with siblings already enrolled in a particular kindergarten are prioritized over children without siblings enrolled, thus children with (older) siblings ‘skip the line’. Fourth, there is a sibling discount with which the (already subsidized) fee is lowered by 30 per cent for a second child and 50 per cent for each subsequent sibling. Quasi-experiments have shown that kindergartens can have a particularly important role in immigrant children’s language formation and subsequent educational outcomes (Drange and Telle, 2015). This difference is likely to be more pronounced for children of immigrant parents than those with native parents. The question of how family policy arrangements influence birth order effects has been raised previously, in particular, whether parents are more ‘effective’ than to day-care in influencing child development (Bonesrønning and Massih, 2011). In the case of immigrant parental ‘effectiveness’, we argue that kindergarten is more effective in native language formation as well as cultural adaptation, and is a possible explanation for the empirical variation documented in this study. Unfortunately, we do not have data on kindergarten attendance. If our assumptions about kindergarten use and birth order patterns are correct, future studies should look into difference in kindergarten use in general, and in particular for whether different immigrant groups display different behaviour in regards to kindergarten use. India, Sri Lanka, and Vietnam Resource dilution theory, confluence theory, strategic parenting, and kindergarten attendance are general mechanisms, whereas we found that only some groups displayed less negative birth order effects compared to those found in native families. There are, however, some more specific arguments for the relevance of these mechanisms in these cases. The three countries that stand out in our study are India, Sri Lanka and Vietnam (abbreviated ISV from here on), as Norwegian-born to immigrant parents from these countries display no negative birth order effects even after controlling for length of residence-related resources at the parental level (via the proxy-variable year of birth at the individual level). Notably, ISV are also the only parental background countries where pupils display school results similar to the native population, as shown in descriptive statistics (Table 1). Norwegian-born to immigrants originating in ISV are known to be high-achievers (Heath, Rothon and Kilpi, 2008; Hermansen, 2016) and have high aspirations for further education. This may be related to a different family-environment and, in turn, different peer-environments for these groups. If parents expect high achievers of all their offspring, they may perform a form of strategic parenting to achieve this, making all siblings perform equally well in the school system. It may also be the case that the language and kindergarten explanations we proposed earlier is only valid for immigrant families with high-achieving expectations, thus making the birth order effects in ISV stand out from the others. Equally, it may be that being Norwegian-born to ISV-parents in Norway is linked to different educational norms, expectations and human capital resources at the peer-group level, as has been proposed for the Indian groups in Norway (Fekjær, 2007). If this is correct, the earlier-born advantage may not be as prevalent in ISV-communities, if the human capital resources has grown with length of residence for the ISV-groups as a whole. It is well known that ISV-communities in Norway organize extracurricular activity regularly, which would counter the within-family mechanism of birth order effects. Limitations Several limitations remain. First, our sample included more second-borns than first-borns, thus introducing a possible between-family comparison in which second-borns (with valid comparisons made to his/her own third-born) from one family are compared to first-borns from a different family. However, sensitivity analyses show that this did not affect our results (see Supplementary Table SA1 and Figure SA1). Second, we have not considered the potential for interactions between age at arrival and birth order, which can be expected to work in opposite directions. Future research should use a sample that includes immigrant children arriving at different ages to investigate this issue in more detail. Third, older siblings who were born abroad before their parents migrated, but who did not themselves migrate are not included in our sample. As our sample includes children who migrated with their parents up until the age of six, any remaining older siblings must be at least 6 years old. Given a typical fertility span of around 20 years for a woman, older siblings may have been left in the origin country between ages of 0–14. We find it unlikely that this practice is very common, and in any case that they influence our results. Fourth, we have argued that there may be cultural variation in birth order effects, as we found variation in birth order effects according to country of origin. However, we did not have direct measurements of individuals’ ethnicities and related cultural variables, but instead relied on information on individuals’ maternal countries of birth. Country of birth is a somewhat imprecise proxy of cultural variation, as in many countries there are many different groups that could have been divided according to other cultural traits, such as religion, ethnicity, and more. This has limited our ability to directly test our claims that birth order effects are partially contingent on cultural variation. Fifth, we did not measure strategic parenting, differences in adaptation to institutions such as kindergarten, or peer effects, which we proposed as explanations. The mechanisms we suggest are theoretical propositions and need further empirical testing. Conclusion In closing, we would like to emphasize some of our key findings and conclusions. We provide what is to the best of our knowledge the first evidence for birth order effects in immigrant populations. There was variation according to parental origin. In most immigrant groups there are birth order effects similar to those in the majority population. Here, the resource dilution and confluence theory are compatible with our findings. In other groups, in particular Norwegian-born to immigrants from ISV, birth order effects are non-existing or positive. Universalistic theories such as the resource dilution theory and confluence theory appears unable to explain this variation in birth order patterns. We have therefore suggested that these theories need to be supplemented by other perspectives that account for cultural and institutional settings, at least in the context of immigrant families. We indicated that early childhood language formation, differences in kindergarten use, culturally rooted strategic parenting, and peer-influences may be potential explanations. Multiple mechanisms may operate simultaneously, however. For example, the processes proposed in the universalistic theories may interact or work in the opposite direction of the other mechanisms that we have discussed. We leave it to further research to address these issues further. Martin Arstad Isungset is a PhD scholar in Sociology at the University of Oslo. His research interests include social science and genetics, social inequality over the life-course, and technological change. Mats Lillehagen is a PhD scholar of Sociology at the University of Oslo. Current research interests include ethnic- and socioeconomic inequalities in the labor market and the family, as well as the demographics of multipartnered fertility and complex families. Elisabeth Ugreninov is a senior researcher at Norwegian Social Research, Oslo Metropolitan University. Current research interests are health-related absence and labour market attachment among ethnic minorities and the disabled. Acknowledgements Data for this article stems from the SEGREGATION project [grant number: RCN project #202479], funded by the Research Council of Norway. The authors thank Torkild Hovde Lyngstad, Gunn Elisabeth Birkelund, Nicolai Topstad Borgen, and the research unit Social Inequalities and Population Dynamics (SIPD) at University of Oslo for their helpful comments. References Aaberge R. , Laake P. ( 1984 ). Om statistiske teoriar for tolkning av data [On theories for statistical interpretation of data] . Tidsskrift for Samfunnsforskning , 25 , 165 – 186 . OpenURL Placeholder Text WorldCat Abafita J. , Kim C.-S. ( 2015 ). Determinants of childrens schooling: the case of Tigray Region, Ethiopia . Education Research and Reviews , 10 , 1130 – 1146 . OpenURL Placeholder Text WorldCat Almond D. , Currie J. ( 2011 ). Killing me softly: the fetal origins hypothesis . Journal of Economic Perspectives , 25 , 153 – 172 . Google Scholar Crossref Search ADS PubMed WorldCat Almond D. , Edlund L. , Milligan K. ( 2013 ). Son preference and the persistence of culture: evidence from South and East Asian immigrants to Canada . Population and Development Review , 39 , 75 – 95 . Google Scholar Crossref Search ADS WorldCat Anh T. S. et al. ( 1998 ). Family size and children’s education in Vietnam . Demography , 35 , 57 – 70 . Google Scholar Crossref Search ADS PubMed WorldCat Barclay K. ( 2014 ). Birth order and intelligence: a within-family analysis using Swedish military conscription data . Stockholm Research Reports in Demography , 14 , 1 – 21 . OpenURL Placeholder Text WorldCat Barclay K. ( 2015 ). Birth order and educational attainment: evidence from fully adopted sibling groups . Intelligence , 48 , 109 – 122 . Google Scholar Crossref Search ADS WorldCat Barclay K. , Hällsten M. , Myrskylä M. ( 2016 ). Birth Order and College Major in Sweden. MPIDR Working Paper. Rostock: Max Planck Institute for Demographic Research. Behrman J. R. , Taubman P. ( 1986 ). Birth order, schooling, and earnings . Journal of Labor Economics , 4 , 121 – 150 . Google Scholar Crossref Search ADS PubMed WorldCat Bertoni M. , Brunello G. ( 2016 ). Later-borns don’t give up: the temporary effects of birth order on european earnings . Demography , 53 , 449 – 470 . Google Scholar Crossref Search ADS PubMed WorldCat Bjerkedal T. et al. ( 2007 ). Intelligence test scores and birth order among young Norwegian men (conscripts) analyzed within and between families . Intelligence , 35 , 503 – 514 . Google Scholar Crossref Search ADS WorldCat Black S. E. , Devereux P. J. , Salvanes K. G. ( 2005 ). The more the merrier? The effect of family size and birth order on children’s education . Quarterly Journal of Economics , 120 , 669 – 700 . OpenURL Placeholder Text WorldCat Black S. E. , Devereux P. J. , Salvanes K. G. ( 2011 ). Older and wiser? Birth order and IQ of young men . CESifo Economic Studies , 57 , 103 – 120 . Google Scholar Crossref Search ADS WorldCat Black S. E. , Devereux P. J. , Salvanes K. G. ( 2015 ). Healthy(?), Wealthy, and Wise: Birth Order and Adult Health. NBER Working Paper (Vol. 21337), available from: http://www.nber.org/papers/w21337 [accessed 22 July 2019]. Black S. E. , Grönqvist E. , Öckert B. ( 2016 ). Born to Lead? The Effect of Birth Order on Non-Cognitive Abilities. IFAU - Institute for Evaluation of Labour Market and Education Policy, Working Paper (10560). Blake J. ( 1981 ). Family size and the quality of children . Demography , 18 , 421 – 442 . Google Scholar Crossref Search ADS PubMed WorldCat Blake J. ( 1986 ). Number of siblings, family background, and the process of educational attainment . Social Biology , 33 , 5 – 21 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Blake J. ( 1989 ). Family Size and Achievement . Berkeley: University of California Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Bonesrønning H. , Massih S. S. ( 2011 ). Birth order effects on young students’ academic achievement . Journal of Socio-Economics , 40 , 824 – 832 . Google Scholar Crossref Search ADS WorldCat Boomsma D. I. et al. ( 2008 ). Intelligence and birth order in boys and girls . Intelligence , 36 , 630 – 634 . Google Scholar Crossref Search ADS WorldCat Booth A. L. , Kee H. J. ( 2009 ). Birth order matters: the effect of family size and birth order on educational attainment . Journal of Population Economics , 22 , 367 – 397 . Google Scholar Crossref Search ADS WorldCat Brekke I. , Mastekaasa A. ( 2008 ). Highly educated immigrants in the Norwegian labour market: permanent disadvantage? Work, Employment and Society , 22 , 507 – 526 . Google Scholar Crossref Search ADS WorldCat Brenøe A. A. , Molitor R. ( 2017 ). Birth order and health of newborns: what can we learn from Danish registry data? Journal of Population Economics , 31 , 363 – 395 . Google Scholar Crossref Search ADS WorldCat Brochmann G. , Kjeldstadli K. ( 2008 ). A History of Immigration: The Case of Norway 900-2000 . Oslo : Universitetsforlaget . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Buckles K. , Kolka S. ( 2014 ). Prenatal investments, breastfeeding, and birth order . Social Science and Medicine , 118 , 66 – 70 . Google Scholar Crossref Search ADS PubMed WorldCat Calimeris L. , Peters C. ( 2017 ). Food for thought: the birth-order effect and resource allocation in Indonesia . Applied Economics , 49 , 5523 – 5534 . Google Scholar Crossref Search ADS WorldCat Conley D. ( 2005 ). The Pecking Order . New York: First Vintage Books Edition . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Dammert A. C. ( 2010 ). Siblings, child labor, and schooling in Nicaragua and Guatemala . Journal of Population Economics , 23 , 199 – 224 . Google Scholar Crossref Search ADS WorldCat Dayioǧlu M. , Kirdar M. G. , Tansel A. ( 2009 ). Impact of sibship size, birth order and sex composition on school enrolment in urban Turkey . Oxford Bulletin of Economics and Statistics , 71 , 399 – 426 . Google Scholar Crossref Search ADS WorldCat de Haan M. ( 2010 ). Birth order, family size and educational attainment . Economics of Education Review , 29 , 576 – 588 . Google Scholar Crossref Search ADS WorldCat De Haan M. , Plug E. , Rosero J. ( 2014 ). Birth order and human capital development evidence from Ecuador . The Journal of Human Resources , 49 , 359 – 392 . Google Scholar Crossref Search ADS WorldCat Drange N. , Telle K. ( 2015 ). Promoting integration of immigrants: effects of free child care on child enrollment and parental employment . Labour Economics , 34 , 26 – 38 . Google Scholar Crossref Search ADS WorldCat Ejrnæs M. , Pörtner C. C. ( 2004 ). Birth order and the intrahousehold allocation of time and education . Review of Economics & Statistics , 86 , 1008 – 1019 . Google Scholar Crossref Search ADS WorldCat Esping-Andersen G. ( 1990 ). The Three Worlds of Welfare Capitalism. Cambridge: Polity Press. Fekjær S. N. ( 2007 ). New differences, old explanations: can educational differences between. Ethnic groups in Norway be explained by social background? Ethnicities , 7 , 367 – 389 . Google Scholar Crossref Search ADS WorldCat Fors H. C. , Lindskog A. ( 2017 ). Within-Family Inequalities in Human Capital Accumulation in India: Birth Order and Gender Effects. Working Paper in Economics No. 700. Gothenburg: University of Gothenburg. Goldstein J. , Lutz W. , Testa M. R. ( 2004 ). The emergence of sub-replacement family size ideals in Europe . Population Research and Policy Review , 22 , 479 – 496 . Google Scholar Crossref Search ADS WorldCat Grätz M. ( 2018 ). Competition in the family: inequality between siblings and the intergenerational transmission of educational advantage . Sociological Science , 5 , 246 – 269 . Google Scholar Crossref Search ADS WorldCat Gualtieri T. , Hicks R. E. ( 1985 ). An immunoreactive theory of selective male affliction . The Behavioral and Brain Sciences , 8 , 427 – 441 . Google Scholar Crossref Search ADS WorldCat Härkönen J. ( 2014 ). Birth order effects on educational attainment and educational transitions in West Germany . European Sociological Review , 30 , 166 – 179 . Google Scholar Crossref Search ADS WorldCat Heath A. F. , Rothon C. , Kilpi E. ( 2008 ). The second generation in Western Europe: education, unemployment, and occupational attainment . Annual Review of Sociology , 34 , 211 – 235 . Google Scholar Crossref Search ADS WorldCat Hermansen A. S. ( 2013 ). Occupational attainment among children of immigrants in Norway: bottlenecks into employment—equal access to advantaged positions? European Sociological Review , 29 , 517 – 534 . Google Scholar Crossref Search ADS WorldCat Hermansen A. S. ( 2016 ). Moving up or falling behind? Intergenerational socioeconomic transmission among children of immigrants in Norway . European Sociological Review , 32 , 675 – 689 . Google Scholar Crossref Search ADS WorldCat Hermansen A. S. ( 2017 ). Age at arrival and life chances among childhood immigrants . Demography , 54 , 201 – 229 . Google Scholar Crossref Search ADS PubMed WorldCat Hoem J. ( 2008 ). The reporting of statistical significance in scientific journals . Demographic Research , 18 , 437 – 442 . Google Scholar Crossref Search ADS WorldCat Hotz V. J. , Pantano J. ( 2015 ). Strategic parenting, birth order, and school performance . Journal of Population Economics , 28 , 911 – 936 . Google Scholar Crossref Search ADS PubMed WorldCat Kautto M. ( 2010 ). The Nordic countries. In Castles F. G. , Leibfried S. , Lewis J. , Obinger H. , Pierson C. (Eds.), The Oxford Handbook of the Welfare State. Oxford : Oxford University Press, pp. 1–25. Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Kavli H. C. ( 2015 ). Adapting to the dual earner family norm? The case of immigrants and immigrant descendants in Norway . Journal of Ethnic and Migration Studies , 41 , 835 – 856 . Google Scholar Crossref Search ADS WorldCat Kristensen P. , Bjerkedal T. ( 2007 ). Explaining the relation between birth order and intelligence . Science , 316 , 1717 . Google Scholar Crossref Search ADS PubMed WorldCat Kristensen P. , Bjerkedal T. ( 2010 ). Educational attainment of 25 year old Norwegians according to birth order and gender . Intelligence , 38 , 123 – 136 . Google Scholar Crossref Search ADS WorldCat Kumar S. ( 2016 ). The effect of birth order on schooling in India . Applied Economics Letters , 23 , 1325 – 1328 . Google Scholar Crossref Search ADS WorldCat Lappegård T. ( 2006 ). Family formation and education among Pakistani, Turkish and Vietnamese women in Norway . Genus , 62 , 75 – 95 . OpenURL Placeholder Text WorldCat Lesthaeghe R. J. ( 2010 ). The unfolding story of transition . Population and Development Review , 36 , 211 – 251 . Google Scholar Crossref Search ADS PubMed WorldCat Lillehagen M. , Lyngstad T. H. ( 2018 ). Immigrant mothers’ preferences for children’s sexes: a register-based study of fertility behaviour in Norway . Population Studies , 72 , 91 – 107 . Google Scholar Crossref Search ADS PubMed WorldCat Lindahl L. ( 2008 ). Do birth order and family size matter for intergenerational income mobility? Evidence from Sweden . Applied Economics , 40 , 2239 – 2257 . Google Scholar Crossref Search ADS WorldCat Lundberg F. et al. ( 2009 ). Maternal smoking during pregnancy and intellectual performance in young adult Swedish male offspring . Paediatric and Perinatal Epidemiology , 24 , 79 – 87 . Google Scholar Crossref Search ADS WorldCat Makino M. ( 2018 ). Birth order and sibling sex composition effects among surviving children in India: enrollment status and test scores . The Developing Economies , 56 , 157 – 196 . Google Scholar Crossref Search ADS WorldCat Monfardini C. , See S. G. ( 2012 ). Birth Order and Child Outcomes: Does Maternal Quality Time Matter? IZA Discussion Paper Series 6825. Bonn: Institute for the Study of Labor. Nadim M. ( 2014 ). Reinterpreting the relation between motherhood and paid work: second-generation immigrant women in Norway . Sociological Review , 62 , 494 – 511 . Google Scholar Crossref Search ADS WorldCat Pavan R. ( 2016 ). On the production of skills and the birth-order effect . Journal of Human Resources , 51 , 699 – 726 . Google Scholar Crossref Search ADS WorldCat Population and Population Change Statistics—Statistics Explained ( 2017 ). Available from: http://ec.europa.eu/eurostat/statistics-explained/index.php/Population_and_population_change_statistics [accessed 16 November 2017]. Price J. ( 2008 ). Parent-child quality time: does birth order matter? The Journal of Human Resources , 43 , 240 – 265 . Google Scholar Crossref Search ADS WorldCat Rodgers J. L. ( 2001 ). What causes birth order-intelligence patterns? The admixture hypothesis, revived . American Psychologist , 56 , 505 – 510 . Google Scholar Crossref Search ADS WorldCat Rodgers J. L. ( 2014 ). Are birth order effects on intelligence really Flynn Effects? Reinterpreting Belmont and Marolla 40 years later . Intelligence , 42 , 128 – 133 . Google Scholar Crossref Search ADS WorldCat Rodgers J. L. et al. ( 2000 ). Resolving the debate over birth order, family size, and intelligence . American Psychologist , 55 , 599 – 612 . Google Scholar Crossref Search ADS PubMed WorldCat Sandnes T. (Ed.) ( 2017 ). Immigrants in Norway 2017. Statistical Analyses , 155, 1 – 190 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Seid Y. , Gurmu S. ( 2015 ). The role of birth order in child labour and schooling . Applied Economics , 47 , 5262 – 5281 . Google Scholar Crossref Search ADS WorldCat Silles M. A. ( 2010 ). The implications of family size and birth order for test scores and behavioral development . Economics of Education Review , 29 , 795 – 803 . Google Scholar Crossref Search ADS WorldCat Smith C. D. , Helgertz J. , Scott K. ( 2016 ). Parents’ years in Sweden and children’s educational performance . IZA Journal of Migration , 5 , 6. Google Scholar Crossref Search ADS WorldCat Sobotka T. ( 2008 ). Overview chapter 7: the rising importance of migrants for childbearing in Europe . Demographic Research , 19 , 225 – 248 . Google Scholar Crossref Search ADS WorldCat Statistics Norway ( 2017 ). Hvordan går det med innvandrere og deres barn i skolen? [How Do Immigrants and Their Children Fare in Schools?], available from: https://www.ssb.no/utdanning/artikler-og-publikasjoner/hvordan-gar-det-med-innvandrere-og-deres-barn-i-skolen [accessed 29 October 2018]. Statistics Norway ( 2018 ). Innvandrere og norskfødte med innvandrerforeldre [Immigrants and Norwegian-Born-to-Immigrant-Parents], available from: https://www.ssb.no/statbank/table/05182/? rxid=71526a61-99ff-48b7-a6ec-7c82838ef7dc [accessed 29 October 2018]. Steelman L. C. et al. ( 2002 ). Reconsidering the effects of sibling configuration: recent advances and challenges . Annual Review of Sociology , 28 , 243 – 269 . Google Scholar Crossref Search ADS WorldCat Steinkellner A. ( 2013 ). Innvandrere og norskfødte med innvandrerforeldre i grunnskolen. En analyse av karakterdata og resultater fra nasjonale prøver i 2012 [Immigrants and Norwegian-Born to Immigrant Parents in Primary School. An Analysis of Grades and National Tests in 2012] (Vol. 2012). Oslo-Kongsvinger: Statistics Norway. Sulloway F. J. ( 1996 ). Born to Rebel: Birth Order, Family Dynamics, and Creative Lives . New York : Pantheon Books . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Sundet J. M. ( 2014 ). The Flynn effect in families: studies of register data on Norwegian military conscripts and their families . Journal of Intelligence , 2 , 106 – 118 . Google Scholar Crossref Search ADS WorldCat Tenikue M. , Verheyden B. ( 2010 ). Birth order and schooling: theory and evidence from twelve sub-Saharan countries . Journal of African Economies , 19 , 459 – 495 . Google Scholar Crossref Search ADS WorldCat Wichman A. L. , Rodgers J. L. , Maccallum R. C. ( 2007 ). Birth order has no effect on intelligence: a reply and extension of previous findings . Personality & Social Psychology Bulletin , 33 , 1195 – 1200 . Google Scholar Crossref Search ADS PubMed WorldCat Wimmer A. , Soehl T. ( 2014 ). Blocked acculturation: cultural heterodoxy among Europe’s immigrants . American Journal of Sociology , 120 , 146 – 186 . Google Scholar Crossref Search ADS WorldCat Workman J. ( 2017 ). Sibling additions, resource dilution, and cognitive development during early childhood . Journal of Marriage and Family , 79 , 462 – 474 . Google Scholar Crossref Search ADS WorldCat Zajonc R. B. ( 1976 ). Family configuration and intelligence . Science , 192 , 227 – 236 . Google Scholar Crossref Search ADS PubMed WorldCat Zajonc R. B. ( 1983 ). Validating the confluence model . Psychological Bulletin , 93 , 457 – 480 . Google Scholar Crossref Search ADS WorldCat Zajonc R. , Markus G. ( 1975 ). Birth order and intellectual development . Psychological Review , 82 , 74 – 88 . Google Scholar Crossref Search ADS WorldCat Zajonc R. B. , Sulloway F. J. ( 2007 ). The confluence model: birth order as a within-family or between-family dynamic? Personality and Social Psychology Bulletin , 33 , 1187 – 1194 . Google Scholar Crossref Search ADS PubMed WorldCat © The Author(s) 2019. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - One Order Fits All? Birth Order and Education in Immigrant Families
JF - European Sociological Review
DO - 10.1093/esr/jcz040
DA - 2020-02-01
UR - https://www.deepdyve.com/lp/oxford-university-press/one-order-fits-all-birth-order-and-education-in-immigrant-families-yFS0ll4VJS
SP - 16
VL - 36
IS - 1
DP - DeepDyve
ER -